Redistribution For Realists


Inequality causes widespread alarm in America.1 Voters list it as one of their major concerns.2 Politicians across the political spectrum describe a “rigged” economy.3 There is a sense that the American Dream is not what it once was.4 What should be done?

The standard economic approach has an answer for nearly all of the law: nothing.

Consider the federal Department of Transportation’s (“DOT”) longstanding practice for distributing grant funds across projects. The DOT relies on cost-benefit analysis to determine where money is best spent. The most important factor in the analysis is the value of time saved. Under current DOT guidelines, saving the time of relatively poor people on buses counts for $25 per hour.5 But, because rich people earn more money per hour, saving the time of richer people at airports counts for well over twice as much: $63 per hour. This practice tends to push transportation spending toward the rich instead of the poor. As a result, many lower-income individuals have difficulty accessing decent jobs, thereby reinforcing existing patterns of income inequality.6

How could it possibly be that baked into our transportation policies is a method that spends more on the well-to-do than on the needy, especially at a time when we are so conscious of the importance of economic mobility?7 Policymakers are not stupid. Some have just come to think that redistribution isn’t part of their job description.

The reason for this restricted approach to redistribution is simple: efficiency.8 Standard economics suggests that policymakers should find the most cost-effective way to redistribute—and then redistribute there and only there.9 And if efficiency is your sole guide for choosing how to redistribute, taxes and transfers typically provide the most benefit at the least cost: Cash lets recipients choose what to spend money on and payers choose what to give up. As leading law and economics scholars Louis Kaplow and Steve Shavell famously argued, the design of nontax “legal rules should be primarily concerned with efficiency rather than the distribution of income.”10 In other words, let tax policy take care of distribution, and all other laws should ignore distributional concerns and instead focus on maximizing the size of the economic pie. Returning to the DOT example, spending more on the rich than the poor maximizes social welfare because the rich can work an extra hour, earning an extra $63, and then an imagined “benevolent dictator” can tax the rich to transfer at least $25 of it to the poor, so that both parties are better off.11

The standard approach makes a lot of sense in theory. It addresses tradeoffs among various options, reflecting the reality of resource constraints. After all, every dollar of government spending can only be used on one thing.

This standard economics approach—described in greater detail in Part II—is prominent among the policy elite that staffs Washington D.C. think tanks,12 advises political candidates and elected officeholders,13 and implements policies such as regulatory cost-benefit analysis.14 Indeed, this thinking is so standard among economic policy elites across the country that it has been called the “Brookings religion.”15 As a result, it has had tremendous real-world influence.16

This Article argues that, while the reigning orthodoxy makes sense in theory, it fails in practice. It ignores how ordinary Americans think about the law and thus ends up exacerbating inequality rather than mitigating it. While elite policymakers support efficient nontax policies that are anti-redistributive, tax policy runs up against political constraints—driven by ordinary people’s attitudes about taxation—that stop tax policy from redistributing enough. This Article plots a different path forward, building foundations for a new law and economics of inequality.

To find a better path, we must pay attention to accumulating evidence on two aspects of the public’s social and political psychology, described in Part III.17

First, people compartmentalize their policy views into silos. They do not think carefully about tradeoffs across different policy areas, and their views on redistribution change with the policy domain. For example, the public may support significant redistribution through transportation policy or other felt “necessities,” like food, housing, and education, while simultaneously opposing redistribution through taxation. The Article introduces the “theory of policy silos” to describe this phenomenon.18 This theory leads to a paradigm shift in the economic approach to law because imagining a benevolent dictator making policy, as if all laws are fungible with all other laws, is no longer sensible.

Second, if one is looking for a domain to realistically expect all redistribution to take place in, tax policy is the wrong choice. Most think the money people earn is in some meaningful sense “theirs” and are thus reluctant to adopt—even hostile to—heavily redistributionist taxation. Partly as a result of this, the Article estimates that the tax code is only about one ninth as redistributive as reasonable baselines suggest that it should be to do all the redistribution needed to maximize welfare.19 The irony, then, is that the one domain where economic theory has insisted all redistribution should take place20 is also one of the places where redistribution proves most difficult.21 The economist’s recipe for reducing inequality turns out to be a recipe for failure, with insufficient redistribution through all means that leaves society far worse off.

As a result, we need to go back to basics and develop a new foundation for law and economics. Since we live in a world with resource constraints, we must take efficiency seriously. However, we must also take data about psychology and political feasibility just as seriously, analyzing policies in the world we live in, not the world as law-and-economics scholars imagine it.22 The Article thus combines the technocratic and the political in ways needed for good policymaking to address inequality in a democracy.

With this foundation, Part IV reveals new law and economics insights on addressing inequality. Most importantly, policymakers typically should redistribute everywhere available—a “thousand points of equity” approach.23 Doing so is actually efficiency-minded. That will give us the most “bang-for-the-buck,” since there are typically at least some low-cost opportunities in each policy area.24 And don’t redistribute too much in one place for the same reason: There are typically diminishing returns to redistribution in any one area. Policymakers should enact an Economic Equity Framework Act that institutionalizes this approach of interpreting the range statutes across the government in ways that promote greater equality, where consistent with other statutory purposes. More generally, a host of institutional considerations arise that have not even been asked, much less answered. These considerations include coordination across different political actors and evolving political attitudes. Part IV begins this project of building a new law and economics of inequality, as does Part V, which discusses critiques.

Crucially, this approach is not just about this particular political moment, with a closely divided Congress. Even without that, it is not reasonable to expect taxes and transfers to do all redistribution. The Article suggests examples from across the law—including regulatory cost-benefit analysis, labor law, antitrust, social insurance, civil penalties, torts, and specific aspects of tax law itself—where such an approach could be implemented.

Of course, some progressives already say things resembling this now. What is lacking is an economic foundation for clear-eyed economic policymaking to address inequality. This Article provides that foundation and thus helps to channel these equitable intuitions in the direction of doing the most good.

Before continuing, I address two features of the Article. First, the Article describes current, but longstanding, attitudes in the United States, without making any claims about “human nature” or other societies, present or future. It also describes general tendencies, not absolute rules. Of course, one Article can only be a partial account, and melding political psychology into law and economics is the beginning, not the end, of a whole new agenda. For example, a general aversion toward high tax rates on the rich to permanently provide cash to help low-income childless adults does not preclude the adoption of temporary modest cash payments as economic stimulus during a global pandemic and depression-level unemployment.25

Second, in making legal prescriptions, the Article assumes the standard economics goal of maximizing welfare by considering both efficiency and fair distribution. Since these goals are important pieces of the picture for many egalitarians, the implications are broadly useful.26 The Article also takes as given the need for further reductions in inequality. However, it cannot say what the exact right distribution of resources is.

Finally, it’s worth noting the precise nature of this paper’s contribution. Needless to say, critique of efficient but inegalitarian lawmaking is longstanding.27 This Article makes two main contributions: one descriptive and one prescriptive.

The descriptive contribution introduces the theory of policy silos and makes a detailed empirical case that depending on taxes alone to achieve distributive justice is not politically realistic. By moving beyond the benevolent dictator to incorporate political psychology, this theory opens up new frontiers in law and economics.

This Article applies this general theory of policy silos to one particular question: how legal rules should redistribute. Much earlier work rejects the standard economics goal of maximizing welfare.28 Of the work that does accept it, other critiques have been offered, like frictions in policymaking (such as a narrowly-divided Senate) or the power of the rich.29 But frictions can be undone over time, and the interests of the rich do not explain why one redistributionist policy in particular does not succeed while others do. In contrast, little attention has been paid to the public’s political psychology.30 As an internal critique, this Article accepts the basic goal of mainstream economic theory but shows how a particular set of durable public attitudes about taxes upends the standard economics approach to inequality and the law.

The contributions of this Article go beyond critique, however. A key question for any legal system is how its various pieces fit together to achieve distributive justice. The Article merges rigorous economic reasoning with political realism in order to articulate a new approach for lawmakers who want to enhance equality but are also attentive to resource constraints and trade-offs, harnessing the powerful tools of economics to achieve equitable ends. The prescriptive analysis provides a new framework for policymakers to address one of the core issues of our time.

The Standard Economic Approach to Inequality:   

The standard economic approach for setting policy considers all policies as a unified whole. This approach is mistaken and produces law that increases inequality. However, before critiquing it, we must first understand it. The approach’s normative aim is to recommend the set of policies that maximize social welfare.31 The Article has no quarrel with this. The standard approach also assumes that policies can be implemented in a coherent and coordinated manner, as if there were a benevolent dictator overseeing all of them.32 This is what the Article takes issue with.

In particular, the standard approach assumes that there is a single economic pie consisting of perfectly commensurable ingredients through which policymakers can and do redistribute, which this Article terms “one-pieism.” Therefore, it makes no difference whether a desired policy outcome is achieved through cash transfers, healthcare, tort law, or other “in-kind” (non-cash) government provisions and regulations.33

Because the “one-pieist” approach considers inputs—the specific policy tools used—fungible, it focuses instead on two key outputs. The first output is the size of the economic pie, with the goal of adopting “efficient” policies that maximize the pie’s size. By doing so, everyone can be made better off, because with a larger pie, there is more to go around. The second output is the division of the pie, with a goal of achieving a fair distribution.

In the one-pieist framework, any adjustments to the initial division should be accomplished through cash transfers and taxes. The standard argument is that taxes and transfers are typically the most efficient means of redistribution.34 Thus, according to the standard economics perspective, everyone is better off when taxes and transfers, as opposed to other policy tools, are used to redistribute.

“One-pieism” ultimately produces the following policy prescription: make all nontax policies as efficient as possible—they should do nothing to reduce inequality—and redistribute only through taxes and cash transfers.35 A key assumption of this approach, to which we will return later, is that it is politically feasible to accomplish all desired redistribution through taxes and transfers.

Efficient Nontax Policies

The one-pieist argument proceeds in two steps. First, make nontax policies efficient. Second, redistribute using taxes. This Section describes the first step: Setting efficient nontax policies to maximize the size of the pie.

By “efficient,” I refer to the “Kaldor-Hicks” efficiency criterion, which maximizes the sum of each individual’s willingness to pay to avoid harm and bring benefits to herself, given the existing distribution of wealth.36 That is, first find what each individual is willing to pay for the adoption of each possible policy. For example, if someone gets a dollar of cash and no one else loses, then the size of the pie increases by one dollar. The same would be true after receiving a good or service that a recipient values at $1. The size of the economic “pie” is the sum of everyone’s willingness to pay for the slice of harms, publicly-provided goods and services, and income that each receives. One can perform this aggregation for each policy option; the policy that fetches the highest amount maximizes efficiency.

Consider again the example of transportation spending. As the Introduction discusses, the Department of Transportation’s longstanding practice for distributing grant funds across projects is to conduct cost-benefit analysis to determine where money is best spent and then spend it there. The most important factor in the analysis is the “value” of time saved. Saving the time of relatively poor people on buses counts for $25 per hour, while saving the time of richer people at airports counts for $63 per hour.37 This practice tends to push transportation spending toward the rich instead of the poor. But it is efficient under the Kaldor-Hicks criterion, because the rich, with their higher wages, are willing to pay more to reduce their commutes.

Suppose that the government has enough money to save one hour of commuting time for some people. It can spend that money exclusively on the rich, exclusively on the poor, or split it between the rich and the poor (helping each group in proportion to the amount of money the group receives). And suppose, for simplicity, that an hour of time savings costs the government the same amount for both rich and poor. Whom should the money be spent on?

The standard economics answer is that the money should all be spent on the rich. That might seem unfair. But the economist’s answer is that doing so makes everyone better off because the rich can work their extra hour, earn an extra $63, and then transfer at least $25 of it to the poor, so that both parties are better off than under an arrangement with equal transportation spending but no transfer. The poor get more than they would have earned if they had worked because of the cash transfer, and the rich have plenty left after the transfer. Put differently, the two parties—rich and poor—should (and, the standard approach presumes, will) strike a trade that makes both better off.38 Both accept an efficient transportation spending rule, and the rich pay more in taxes to fund transfers to the poor.

This “law and economics two-step” has a more common lay description: Maximize the size of the pie and then divide it up equitably. One sees such reasoning, for example, in the context of international trade: We should have free trade, maximize the size of the pie, and then divide up the pie to achieve a fair distribution.39

Importantly, “redistribution” is a term of art in economics. It does not mean to distribute more to the poor than today or more than seems fair. Instead, the reference point is the efficient amount. So, if the efficient amount of transportation spending on the poor is $10 million, but the government spends $15 million, that counts as $5 million—not $15 million—of redistribution. If the government is spending less than $10 million today, the government should spend the full $10 million on the poor because it would benefit both efficiency and equity. The question here is whether the government should spend beyond $10 million on the poor.

That is, efficiency reasons may still dictate government provision of nontax benefits to the poor. For example, in reality, there will be some low-cost, high-time-saving investments for the poor, so they will get some funds for reasons other than “redistribution.” But the spending is heavily tilted toward the rich, since their time counts for over twice as much. (See the Appendix for a formal model and numerical simulation.) Similarly, information problems in insurance markets can justify a significant role for the government in healthcare provision.40 And sometimes, it can be efficient to provide in-kind benefits (like low-end housing) rather than cash benefits to “target” the truly poor, who would be especially willing to accept the in-kind benefits.41 There are many such arguments. But these noncash benefits all have one thing in common: The reason for providing them is efficiency.

The standard story thus reduces policymaking analysis to the consideration of a single “pie” with two essential characteristics: size and distribution. The views of the general population about how the pie grows or is distributed (e.g., healthcare and environmental health versus bananas and automobiles) play no role.

Redistribution Through Income Taxes

This Section explains the second element of “one-pieism”: redistribution. The standard economics approach cares about distribution, too. But according to the standard approach, redistribution should take place through taxes and transfers, not other means. The key intuition for redistributing with cash through taxation rather than redistributing with in-kind goods, services, or legal rules is that cash lets the poor choose what to spend the money on. Using cash thus increases the likelihood that resources are used in the way that makes them best off. By the same logic, decreasing the cash income of the rich through taxes, rather than removing some legal entitlement the rich may highly value, aims to harm the rich the least and thereby enables greater redistribution to the poor. Given that the government has limited resources and thus faces trade-offs in how it addresses equality, cash—and thus taxes—is viewed as the superior means of redistribution. But how should those taxes and transfers redistribute?

Returning to the example of transportation spending, few are so unrealistic as to imagine that the DOT itself would make cash transfers to the poor when it spends on projects that favor the rich. It lacks such legal authority. Instead, the two-step process imagines that the cash will typically be transferred through the income tax and transfer system, in ways not specifically tied to transportation spending but rather in ways that will yield roughly fair distributions in the end.42         Maximizing the size of the economic pie is not the ultimate goal of taxes; the distribution of the pie matters for welfare, even if it means shrinking the size of the pie.43 So, for decades economists have studied how taxes should redistribute. Distribution matters because the welfare that a rich person gets from an additional dollar of consumption is generally less than the welfare that a poor person would get from that additional dollar.44 Thus, maximizing overall welfare across everyone often requires transferring resources from the rich to the poor. The desire to redistribute to the poor could also be grounded in broader equality concerns shared by a variety of theories of justice. So one need not accept the specific normative underpinnings of optimal tax theory to value its recommendations on how to set redistributive tax policy.45

Two key elements of the standard economic approach on taxation concern us at this point.

First, the approach assumes that all “needed” redistribution takes place through taxation. The overall amount of redistribution can be determined in a variety of ways. That goal could come from external sources, including the declining marginal utility of income, the distribution of welfare across individuals, and interdependent preferences for welfare.46 But typically it is assumed that members of the public have overall distributional preferences that the government should follow.

Then we imagine that taxes redistribute to achieve that distributional goal, while considering efficiency losses. If someone is taxed to transfer resources to the poor, they may work less hard, shrinking the size of the pie, like crumbs falling as pie is transferred from one person to another.47 As a result, standard economics does not suggest that redistribution should result in equalization—in slices of pie of equal size. Indeed, this tradeoff between efficiency (the size of the pie) and equity (fairly redistributing) is at the center of economics. But a key point of this weighing is that, whatever the redistributive goal is, taxes are assumed to fill all of it, on their own, without receiving help from other policies.

It follows from this imagined role of taxation that, for prescribing tax rates, it does not matter who earns the money, except for the incentives. In other words, making the money yourself gives you no moral entitlement at all to keep it. Of course, there may be incentive-based reasons to not tax income away. But the theory implies that people are not entitled to any of their income because those taxes need to achieve overall distributive goals, which are not impacted by who happens to initially earn a dollar of income.

Second (and as an implication of the first element), taxes redistribute fungibly with other means of redistribution. That is, everything that poorer individuals do not receive through in-kind government provision—like roads or other efficient (or inefficient) redistribution—will happen through taxes. So, if the government cuts back on an inefficient in-kind provision of transportation worth $1 to the poor, it assumed that the poor will receive at least $1 through taxes and transfers. For the purposes of how redistribution should and does happen, it does not matter that transportation and taxes are different forms of redistribution. A dollar of value in different forms is still a dollar. And the benevolent dictator will adjust taxes appropriately to fulfill the distributional goals.

 Note that these taxes are imposed on labor (also called “earned”) income. Capital income taxes have traditionally been viewed as inefficient, in part because capital is so mobile, and thus a bad means of redistributing.48 These optimal labor income taxes are the taxes and transfers that the prescription of “efficient nontax policies” depends upon.49 So, when the Article discusses public attitudes about taxes, it means the taxation of labor income.

This so-called “optimal tax theory” dates to the 1970s.50 At that time, well before the advent of behavioral economics, it was standard in economics to assume rational, self-interested individuals with simple tastes. Optimal tax theory assumes that individuals want to consume more and that they do not care about how redistribution should take place.51 This theory has played an important role in the development of modern law and economics by providing a (supposedly) clean and efficient tool for achieving desired distributional outcomes.52

Part III describes more specific implications of optimal tax theory. It shows that this theory prescribes some specific and, in some cases, surprising, things. These things have not come to pass in practice, perhaps because the theory ignores the behavioral insights that have revolutionized many parts of economics—but has done little to touch the law and economics of inequality, until now.

Legal Impacts

In contrast, the efficiency part of the standard economics prescription—the first prong—has had huge impacts on many areas of the law. It has reshaped diverse swathes of legal rules to become more efficient but less egalitarian. Consider two examples: tort law and civil procedure.

Tort law has been strongly influenced by the national movement for tort reform, which emphasizes a goal of efficiency.53 The movement has “foster[ed] efficient behavior” by, among other things, ensuring that “damages awarded do not exceed the amount necessary to provide ‘reasonable compensation,’”54 even though such changes tend to benefit (often richer) defendants over (often poorer) plaintiffs.55 To achieve its goals, the tort reform movement pursues “pro-defense changes in tort law,” including reducing the amount of damages available by altering “tort doctrines[] such as . . . joint and several liability, collateral sources, punitive damages, and noneconomic damages.”56

Shifts in civil procedure also embody the law and economics emphasis on efficiency over redistribution. Take, for example, the shift toward arbitration over trials. Access to a full trial and to jury adjudication of civil wrongs has long been an important method for redistributing wealth away from defendants (who are, again, often corporate and richer) to plaintiffs (who are, again, often lower income), partly because of sympathy by juries “to the claims of a consumer [over] a large company.”57 But the law and economics quest for efficiency is slowly shutting off these traditional methods of access to the court system. For example, in AT&T Mobility v. Concepcion,58 which strengthened the role of arbitration, the Supreme Court majority “began by extolling the virtues of arbitration over court adjudication in terms of efficiency” and discussing “the ill effects of class actions on businesses.”59

Scholars describe similar trajectories in many other areas of law.60 For example, since the 1980s, major federal regulations have had to go through efficiency-based cost-benefit analysis, overseen by the White House Office of Information and Regulatory Affairs.61 Corporate law adopted a goal of shareholder-value maximization.62 And, under the influence of scholars like Robert Bork,63 antitrust law moved from enacting broader notions of fairness to a focus on economic efficiency.64 In all of these fields, equality has been subordinated to efficiency.

The impact of the efficiency approach—the first prong of one-pieism—has been tremendous in the real world. As we will soon see, the second prong—redistributionist taxation to address distributive concerns—has had less of an impact in the real world. This is perhaps in part because tax policy is typically set in salient political battles in Congress, whereas in other policy areas experts who are less responsive to voters and more responsive to economic education hold more sway.65 This asymmetry is key to the failure of the standard economic approach.

Of course, the widespread adoption of efficient but inegalitarian legal rules under the influence of law and economics has not gone uncritiqued. Outside of economics, many have criticized economic policy prescriptions.66 Political scientists have proposed alternatives, such as “pre-distribution,” that would redistribute “through the market,” though without a specific account of what counts as pre-distribution, why it is superior to the economics prescription, or how to redistribute that way.67

This Article provides a specific account of what is wrong with the standard approach and a program for what to do about it, based on recent insights from psychology and economics. The Article rigorously considers the social reality of income taxation, which is at the crux of the law and economics arguments, shows how this reality transforms economic policy prescriptions, and develops a new approach given this reality.                                                                       

The Standard Approach Fails and Exacerbates

The standard one-pieist economic theory of having all redistribution take place through taxes does not work in reality. By relying solely on taxation, it leads to too much inequality. As this Part details, redistribution through taxation is strikingly limited. Indeed, new estimates here show that the tax code is only about one ninth as redistributive as one baseline suggests that it should be.68 This is a large part of why inequality is so high.

Two key factors from commonplace social and political psychology help explain why. First, the public siloes its attitudes about policies: Rather than thinking holistically about all policies, the public tends to think about policies in isolation and apply different attitudes to different policies. This Article introduces the “theory of policy silos” to explain these public attitudes. The key result is that people are more comfortable redistributing through some means than others.

Second, the public resists redistribution through the tax policy silo. To ordinary people, fairness in tax does not mean “optimally redistribute from rich to poor.” While there could be a variety of reasons for the resistance to redistribution in the particular silo through which redistribution is supposed to take place—taxation—the Article explores one reason in depth.

In particular, to a significant extent, the public believes that people deserve to keep their income. They believe they deserve to keep it because they worked for it and own it, which makes it harder to tax away. For the standard economic prescriptions to work, with income taxed and transferred away to produce a fair distribution, the public must believe that—apart from producing good incentives—your money is as much mine as yours. But the public does not believe that, as this Part shows much evidence to support.

As a result, taxation is insufficiently redistributive, and an approach of adopting efficient but inegalitarian nontax policies increases inequality and harms well-being. Consequently, the standard economics approach should be rejected.

The Public Siloes Its Policy Attitudes

Most ordinary people do not view policy like one-pieist economists do, imagining a benevolent dictator trading off one policy against another to find the most efficient way to redistribute. Rather, according to the theory of policy silos introduced by this Article, ordinary people hold category-by-category views about what is just for a given policy and apply those views partly in isolation. Each policy is its own “policy silo.” Attitudes about redistribution apply differently across silos. These attitudes can vary based on the actor implementing the policy or the good or service impacted.

In short, the dominant economic metaphor is mistaken. In terms of what happens in the real-world policymaking, there is not one pie. Instead, there are many silos—some silos for which the public is comfortable having a high level of redistribution and others less so. And, crucially, individuals’ support for redistribution can differ depending upon whether they are considering taxation or other means of redistribution.

The theory of policy silos builds in part on behavioral economics work on “mental accounts.” The example typically given for mental accounting is savings behavior.69 Standard economics assumes that money is fungible: People should treat a dollar the same way regardless of how they receive it.70 But people do not act that way. For example, even though a dollar is basically the same whether it is earned through one’s wages, appreciation in stock, or another means, the likelihood of people increasing their consumption in response to earning an additional dollar is often different depending upon which “mental account” that dollar falls into. Economics Nobel Laureate Richard Thaler’s explanation is that people adopt prudent rules-of-thumb in their financial decision-making. For example, believing that one should “live within your means,” people consume out of their current wage income much more than out of increases in the value of their stock.71

The theory of policy silos holds that, in many ways, individuals view policy in the same way that they view their personal mental accounts: with different attitudes applied to each silo. Economists tend to view welfare maximization as beginning with one big pie, containing taxes, environmental health, healthcare, education, transportation spending, minimum wage laws, etc., that can be distributed with perfect commensurability across policies as the social planner sees fit. This Article suggests that many people, instead, view distributional issues on a category-by-category basis, like they view their personal finances. Of particular importance for this Article is that people have policy views on taxes, which are often not only about proper incentives and redistribution but also about people deserving to keep the money that they earn.

This theory may simultaneously be intuitive to some readers—which is good, as the argument is that this is commonplace psychology—and also upend standard policy prescriptions. For example, people may think that it is good policy to spend on transportation to the poor, thus “redistributing” to them, so that they have the opportunity to get to work, but oppose “giving out” cash transfers through the tax system. As well, people think about transportation and taxation at least partially separately, not as part of one big fully fungible system.

This Article focuses on the tax policy silo, which is described at length in the next Section. A description of all potential policy silos is beyond the scope of the Article. But, to illustrate distributional attitudes in one policy silo that are at variance with those imagined in the standard approach, consider government provision of in-kind “necessities.” The standard approach holds that the government should not redistribute through provision of necessities beyond the “efficient” amount.72 Nevertheless, support for redistribution through necessities appears widespread.

There are various indications of how redistributive attitudes are particular to the silo for necessities. For one, rhetoric regarding necessities is often about rights. For example, UN Declaration of Human Rights delineates in-kind rights “including food, clothing, housing and medical care.”73 It does not say that people should get cash to spend as they wish. Even in the United States, many seem to believe in a “right” to healthcare based on polls.74 And across countries, there is widespread in-kind redistribution in areas considered necessities, such as healthcare, education, housing, food, and childcare.75 Some of that spending can be justified on standard efficiency grounds, such as high returns to education for children who cannot pay for it themselves.76 But not all spending can be justified that way. For example, there is lots of redistribution through housing even to the elderly, who could presumably choose whether to spend cash on housing or other things absent a behavioral failing.77 And the United States likely provides healthcare to people that costs far in excess of what poor people would be willing to pay for it.78 This policy exists despite the traditional efficiency analysis suggesting an improvement that would make everyone better off: providing those people unwilling to pay for a full complement of healthcare with (lower value) cash transfers through the tax system.

One could consider many other separate policy silos as well.79 For example, people seem unwilling to tolerate valuing the lives of the rich more highly than the lives of the poor for the purpose of regulatory cost-benefit analysis—which would be required in any efficient regime, since the rich are “willing to pay” more for their lives owing to their greater ability to pay.80 Other nontax policy silos may be indifferent to the needs of the poor. For example, in the context of torts for property damage, the goals may be primarily compensatory or to deter bad behavior, rather than to address distributional impacts between the rich and poor.81 It seems clear that people hold category-specific redistributive attitudes. And just because an individual gets more in one category does not mean that people think that she should get less in another category. That’s how economists think, but often not how non-economists think. These category-specific attitudes about fairness have major implications for the right policy to adopt in a given real-world situation across economic questions in the law.

Ample evidence from survey experiments supports the idea that people compartmentalize policies. Consider a few examples. First, new evidence from one survey experiment shows that support for programs that seem identical to economists have considerably different support based on how they are administered.82 The study compares support for government programs to fund individuals’ expenditures, such as rent. Respondents were randomly asked about their support for programs to help cover rent in one of two ways: through reducing taxes by $1,000 or through having a separately administered spending program. Although the two programs would be considered by economists to be essentially identical, since $1,000 in reduced taxes should be the same as getting a $1,000 check in the mail, support is considerably higher for the program reducing taxes than the program increasing spending.83 The tax silo appears different from the government spending silo; letting people keep their pretax income appears fairer than taxing it away and then having the government spend it.

Similarly, a second study shows a wide gulf between support for a “cash” transfer and a transfer to be spent on “necessities.”84 A large, demographically representative group of Americans were randomly presented with one of two identical scenarios with a $2,000 transfer to the poor.85 In one scenario, the money could be spent on anything.86 In the other, “necessities account” scenario, the money could be spent only on healthcare, food, and housing.87 Support for this “necessities account” was a gaping 29 percentage points higher after being presented with both options.88 The difference was even larger for the richest, possibly most politically influential,89 respondents. Respondents were asked how they think about this issue, and the most common responses were about ensuring that people buy “basic necessities” with “taxpayer” money. Necessities appear to be a different silo than cash—and one in which government redistribution seems far more appropriate.

People even compartmentalize between different types of taxation, a third study shows.90 One experiment considered how support for taxation varied depending on whether it was split between two different forms of taxation: A tax on incomes and a separate tax on “payroll,” which was effectively the same as a tax on income in the context. Respondents preferred a considerably higher total amount of taxation when both forms of taxes were present—on income as well as payroll—than when only one was present.91 The payroll tax and the income tax seem to be in different siloes.

People could have policy silos for a variety of reasons, including biases and legitimate values. The third study, on preferring higher total taxes when both income and “payroll” taxes are enacted, looks more like a bias of failing to aggregate across two types of taxes. Even in personal financial decision-making, with huge individual stakes, people tend to fail to aggregate across financial accounts. So it is unsurprising that in their thinking about policy issues, where the personal stakes of one’s actions like voting are trivial by comparison, they would make similar errors.

The second study, on preferring necessities accounts rather than cash, may look more like a value about certain necessities being good to spend money on or at least good for a government to provide.92 Or as the survey experiment suggests, it could also partly reflect confusion about how poor people spend their money, with many people paternalistically—and incorrectly—believing that the poor will spend a large share of their funds on things other than “necessities.”93

The first study, showing a preference for reducing taxes over increasing government spending, could be a bias, a value, or something else. In other cases, voters may consider policy-by-policy fairness to more easily monitor politicians. There is a huge amount to unpack in the theory of silos, with potential normative and practical implications across law and economics. But those implications—outside of redistribution—are mostly left to future work: Law and economics scholars should develop the vast swath of policy implications arising from policy siloing. The key point for this Article is only that policy siloing about redistribution is an important feature of reality.

Public Attitudes Hinder Redistribution Through

This Article focuses on one particular policy silo—taxation of labor income—because such taxes are supposed to be the means of redistributing. Public attitudes substantially hinder redistribution through taxation. There could be many reasons, such as greater resistance because of the high salience of the one big policy of taxation versus the lower salience of many smaller policy issues. This Section focuses on one reason: In the tax silo, many people care about desert based on pretax income. That is, many believe that if people earn more money, they deserve to keep a decent share of it, and, if they earn less, they deserve less money, all for reasons unrelated to incentives to work.94

This commonplace psychology is inconsistent with the standard economic approach’s tax policy prescription of adopting all redistribution through taxation. With partially siloed policy views, it is completely consistent to resist redistribution through the specific process of taxes, while at the same time supporting either more equal outcomes overall or more egalitarian provisions through specific goods (like healthcare) or processes (like regulatory cost-benefit analysis).

Policy silos are a key framework for understanding public attitudes because desert, and a host of other concerns, may enter each policy silo, but may apply differently. For example, people may deserve to keep their incomes, limiting redistribution through one means, but they may also deserve to receive support from the government for certain necessities, increasing redistribution through another means. Ignoring this reality dooms the standard economic approach to make poor policy prescriptions.

Four sources suggest that notions of desert in pretax income that limit tax-based redistribution are important to how people think: (1) philosophical reflection; (2) the large gulf between the real world and the standard economics prescriptions for taxation; (3) survey experiments; and (4) political discourse. This Section reviews each in turn.

Philosophical Reflection and Introspection

Turning first to philosophical reflection and introspection, the seminal statement on tax desert is Liam Murphy and Thomas Nagel’s The Myth of Ownership: Taxes and Justice.95 They argue that the belief that people deserve to keep a substantial share of their pretax income comes from two sources: the senses that people own and earn their pretax income.96 Ownership can seem to confer a sense of “unqualified moral entitlement to what we earn in the market.”97 And the idea “that higher market returns are in some sense deserved as a reward arise[s] naturally within the everyday outlook of participants in a capitalist economy.”98 Desert is then strengthened by the sense that people not only own their pretax income but also earn it, either from thrift in savings or from industrious hard work or skill. This belief then easily “slides into the much broader notion that all of pretax income can be regarded as a reward for those virtues.”99

Murphy and Nagel argue that desert in pretax income is mistaken because pretax income is largely determined by legal structures for which individuals have little direct responsibility.100 Nevertheless, “banish[ing these ideas] from our everyday thinking” would be difficult101 and “counterintuitive [because t]axes are naturally perceived by most people as expropriations of their property.”102 It would be difficult for the idea that income is not “theirs” to “become psychologically real to most people [because p]retax economic transactions are so salient in our lives.”103 Instead, what is psychologically real “is the robust and compelling fantasy that we earn our income and the government takes some of it away from us, or in some cases supplements it with what it has taken from others.”104 They suggest that “[c]hanging this habit of thought would require a kind of gestalt shift, and it may be unrealistic to hope that such a shift in perception could easily become widespread.”105 The claim is not that taxation is impossible but that such views lead to a drag on the ability of the state to redistribute through taxation.

The commonplace attitudes that Murphy and Nagel describe are inconsistent with the standard economics approach. That approach imagines all resources combined into one pie and then reallocated to maximize welfare. It does not matter who has what to begin with; there are no rights, and there is no desert based on pretax income. Individuals are completely dissolved in optimal tax theory, except as producers of income facing incentives and then as consumers of income, untied to who produced what. Of course, allowing people to keep part of their income is important to encourage work. But desert-based attitudes are different. With such attitudes, pretax income generates desert. As a result, a more unequal pretax income distribution will tend to increase the desired inequality in post-tax incomes, even ignoring incentives.

Failure of Existing Institutions to Follow Optimal Tax

Existing institutions provide additional evidence of the dissonance between reality and what taxes would actually need to do under the standard approach. Economists have now extensively studied the tax rates required to achieve a welfare-maximizing distribution of income, given the behavioral response to taxation.106 But this optimal taxation literature has received little notice in the legal literature outside of tax scholars themselves.107 Of course, there is no exact agreement on what an “optimal” income tax would look like.108 However, there is broad agreement on several features that maximize the benefits and minimize the costs of using taxation as a redistributive tool.109 Wherever tax policy is used for redistributive ends, these features are likely to appear. However, the magnitudes will differ depending upon the desired distributional outcome.

 This Article is the first to carefully juxtapose the high level of redistribution predicted by optimal tax theory with the substantially more modest redistribution we see in reality, and draw out implications for the law outside of taxation. Existing redistribution is telling about commonplace views on taxation, assuming that the real-world evolution of political institutions reflects those commitments to some extent. This Subsection reviews five important prescriptions of optimal tax theory, none of which are borne out in practice. It then concludes by comparing the redistributive views implicit in the tax code with the redistributive views people tend to hold about their own incomes.

First, optimal tax theory typically prescribes a large cash “demogrant” (essentially, a universal basic income) to let people choose what to spend their money on.110 This demogrant goes by many names but has received support from thinkers on the left and the right. For example, in the 1970s, Richard Nixon and Milton Friedman advocated for a “negative income tax,” essentially a universal basic income by a different name.111 Though the exact size that would maximize overall welfare is unclear, estimates are generally in the range of several thousand dollars annually. Most recently, a sophisticated estimation by economist Emmanuel Saez recommends an annual demogrant of $11,900 (in 2018 dollars).112

Without denying that some policies give some cash to some people (especially children, who are unable to work), the government provides no such demogrant or anything close to it.113 For example, a childless adult who is neither working nor in training may be eligible for little or no cash support from the government.114 This outcome is consistent with the importance of desert and a resulting distaste for unrestricted cash transfers: The state typically does not give out money to those who do not work for it, without some strings attached.115

Second, the demogrant is taxed away at high marginal tax rates for those with low incomes. For example, Saez’s model on average has 60 percent marginal tax rates on those earning between $9,800 and $24,500 (in 2018 dollars).116 This maximizes tax revenue (since lots of taxpayers earn at least modest amounts of income) while distorting the behavior of relatively few taxpayers (since many taxpayers will earn well beyond that amount in any case).117 Nevertheless, this creates a large disincentive to work for lower-income taxpayers. Saez estimates a resulting unemployment rate of 13.8 percent because of the large demogrant and high taxes.118 The disincentive to work for lower-income earners is considered a worthwhile efficiency cost because of the taxes collected on higher-income earners.

Again, we do not see high tax rates on modest incomes. Rather, it seems unfair to have high tax rates because workers would not be getting a fair share of output of their labor. Indeed, those high tax rates would discourage large numbers of people from working at all, eliminating their workplace contribution to the state altogether.

Third, fixed attributes of people should be “tagged”119 to observable characteristics correlated with earnings ability.120 That is, for two taxpayers earning the same income, the one with the characteristics that are correlated with having the higher earnings potential, like height, should be taxed more.121 While taxing income incentivizes people to work less, taxing based on fixed characteristics that are correlated with earnings ability reduces the efficiency loss. It still partially taxes those who can earn more, but because fixed characteristics cannot be changed, it does not discourage work. In the extreme, if the government were omniscient, it could tax based on its knowledge of an individual’s earnings ability and not cause any distortion at all, since people would not be able to reduce their tax burden by changing their behavior.122 Additionally, since the highly-skilled are so much more productive, they should pay high taxes, possibly so high that they need to work long hours in jobs that they dislike to provide more resources to be redistributed.123 That is, since the costs of work are presumably similar for everyone but the benefits for society of work are so much higher for the high-skilled, the high-skilled should effectively be forced to work much more.

In practice, though we do see some tags like disability, we do not see taxation based on tags like height.124 Nor does a tax based on height seem like a plausible policy to adopt, since taxing people who earn the same income based on height may simply seem unfair. Evidence from survey experiments confirms that very few people support such tags.125 This result is also consistent with an aversion to setting taxes on the basis only of incentives and redistribution, rather than also internal fairness norms. If two people produce the same amount, they should be treated the same by the state, even if one is tall and one is short.

Fourth, optimal tax theory says that, as inequality increases, taxes should become more redistributive.126 That is, as the share of income earned by those with the highest incomes goes up, their tax rates should go up.127 This makes sense: As the rich get richer, their marginal utility of consumption declines, making it worth taxing them more.

Yet, despite the well-documented rise in income inequality,128 there is widespread agreement that taxes have not kept up.129 This failure could be due in part to the strikingly low support for high taxes on the rich. Recent surveys show only roughly half of Americans want high taxes on the rich.130 The large numbers of relatively poor people who do not want higher taxes on the rich is especially striking, with 26 percent earning less than $25,000 opposing higher taxes on the rich and an additional 17 percent neutral.131 This result is also consistent with the importance of desert: people get what they deserve. And, if inequality goes up, that does not necessarily mean that the rich should pay much more because their skill and work effort produce pretax income, of which people deserve to keep a large share.132

Fifth and relatedly, taxes should roughly be set to maximize revenue from the rich. With a declining marginal utility of income, at very high incomes, the benefit of an additional dollar is essentially zero.133 Thus, taxes should be set such that the government can collect as much as possible, setting a quite high rate that discourages work somewhat, but not to the point of reducing tax revenue. In other words, rich individuals are treated “only as the revenue producing property of the state.”134

We do not see such high tax rates, perhaps because (absent a war or some perceived great social need, recent historical research suggests135) people believe that well-off individuals—as individuals who have large talents and work hard—deserve to keep a decent share of their income. And again, the theory of policy silos shows that none of this is inconsistent with the public’s desire for more redistribution in general or in specific nontax places.

*   *   *

Recent advances allow economists to quantify and summarize the redistributive views implicit in the tax code, by extrapolating the value we must place on income in the hands of the poor to be willing to distort behavior with taxes and transfers to help them.136 Nathaniel Hendren shows that our current tax code implicitly values a dollar in the hands of someone at the tenth percentile of income ($14,000) 1.5 times as much as someone at the ninetieth percentile of income ($179,000).137

By contrast, individuals’ own personal behavior regarding risk implies a far higher weight on the poor relative to the rich.138 For themselves, they value an extra dollar about thirteen139 times more highly if they had an income at the 10th percentile than if they had an income at the ninetieth percentile.140 The reason is that people value the things (often necessities) they would buy at low incomes more than the things (often luxuries) they would buy at high incomes. The results come from looking at, for example, how much people are willing to pay for insurance, which essentially redistributes money to times when the individual is poorer because of loss of an asset like a house burning down.

So, our current tax code redistributes across people nowhere remotely close to the amount that individuals do internally for themselves. Given the implicit ratio of values between the tenth and ninetieth percentile income-earners of 1.5 in the tax code versus 13 internal to individuals, the tax code implicitly only gives poor individuals about one ninth of the weight implied by people’s own behavior. Of course, all may not agree that this method—using individuals’ own behavior—is the right way to measure welfare.141 But, it would have to be wildly far off for taxes to be anywhere close to redistributing the right amount.

Inferring commonplace views from existing policies while trying to offer recommendations to improve policy leads to an obvious circularity problem.142 Policies could look the way they do for many reasons, including political capture.143 Nevertheless, the difference between what taxes would probably need to do under optimal tax theory and what they in fact do is remarkably gaping. This is especially so because there are few technological barriers to any of the implications of optimal taxation. We have simply chosen not to do those things. And, in any case, as the previous Subsection explained, another account would need to explain the great deal of redistribution through nontax means. Maybe asking the tax code to provide the redistribution necessary for a just society is simply asking ordinary people to accept something inconsistent with their moral intuitions.

Empirical Evidence in Psychology and Economics

Emerging empirical evidence in psychology and economics further shows how public attitudes support limiting tax-based redistribution in ways inconsistent with one-pieist goals of maximizing the size of the pie and dividing it up to achieve a fair distribution. The studies work by isolating how people reason about distributional concerns in tax policy. While there have now been several survey experiments on the topic,144 consider two illustrative ones.

One study shows the widespread perception that people deserve to keep what they earn, even based on genetic advantages.145 Christopher Freiman and Shaun Nichols present respondents with a case of two jazz singers, each of whom works equally hard, but one of whom is a better singer because of genetic differences. Respondents said that the better singer deserved more money and that it was fair for her to receive it. This empirical result conflicts with the prescription from optimal tax theory that both singers should receive the same post-tax income, apart from incentive considerations—which are not at play here because these skills are inborn.146 Instead, this evidence suggests that even pretax income arising from inborn skills generates perceptions of desert.

In another experiment, Matthew Weinzierl shows the prevalence of beliefs that desert inheres in pretax income, even if earned for reasons unrelated to effort.147 Survey respondents are presented a hypothetical in which two people are assigned arbitrary pretax incomes, one richer and one poorer.148 Respondents are then asked how much tax each party should pay to help fund a specified amount for necessary public goods, as well as potentially redistribute across the two parties.149 This setup removes incentive effects of taxation, since the parties receive the money irrespective of behavior. As a result, the standard economics approach prescribes equalizing the incomes of the two people because a dollar continues to produce more utility for the poorer person until the incomes are equalized, and higher taxes will not shrink the total income available to the two parties. Nevertheless, a large majority of respondents (75 percent) stop short of full equalization, and many stop well short.150 Put differently, the entirely arbitrary “pretax income” appeared to have moral weight. Pretax incomes appear to generate perceptions of desert.

Nothing in the experiment suggests that people are not also motivated in part by standard goals of achieving fairer outcomes; respondents preferred some redistribution. Nevertheless, this evidence suggests that people have some attitudes quite distinct from the one-pieist redistributionist ones that economists typically focus on. Pretax income—even if entirely arbitrary, arising from factors like natural ability over which people have no control—appears to drive many people’s views of fair taxation.

Summarizing the empirical literature on attitudes about redistribution in taxation, Steven Sheffrin writes in Tax Fairness and Folk Justice that ordinary people are often less concerned with “distributional issues, or ‘who gets what’” than of having a fair method linking pretax incomes and post-tax outcomes.151 Indeed, new evidence shows that a whopping 80 percent of Americans agree that, “[p]eople deserve to keep the money they earn.”152 Because many people believe a fair method requires connecting a person’s ultimate distributional outcome to what he initially earned, they tend to resist the redistribution imagined under one-pieism for achieving distributional objectives purely through taxation.

Political Discourse

Fourth, our political discourse—which presumably is in part designed to appeal to people’s basic intuitions—suggests that many people believe that desert inheres in pretax income. For example, President George W. Bush often described tax cuts as letting Americans “keep more of their hard-earned dollars,” an apparent appeal to hard work generating desert.153 While Bush mentioned incentives too, the issue of distributive justice—setting tax rates based on need, a key factor in optimal taxation—was often entirely absent. The liberal crop of 2020 Democratic presidential candidates seemed to at least partially agree. With the exception of Andrew Yang (who polled in the low single digits and got 0.42 percent of the primary vote),154 even the most liberal candidates did not propose providing large unrestricted cash transfers to the poor. And Democrats commonly use similar language about desert.155 For example, in response to the question, “Does anyone deserve to have a billion dollars?” then-Senator Kamala Harris responded, “If they earn it and work hard for it, sure.”156


In short, one of the two implications for the law and economics two-step is imposing an “optimal” tax. Yet people appear to hold attitudes at odds with the implications of optimal taxation, and current policy is far off from those implications. Scholars disagree about whether desert should inhere in pretax income.157 But there is strong evidence that this attitude is commonplace, which is the only claim this Article makes. What has been missed in the debate thus far—which has focused on the implications of these attitudes for taxation—is the implication of the apparently durable commonplace attitudes for a vast array of nontax policies. To be sure, views can change, but the standard approach makes very strong assumptions about what taxation will do, and there is little reason to premise all nontax policy on these assumptions given available evidence.158

Overall, the Article operates from a theory about the social and political psychology of voters: Many have, in significant part, desert-based views about taxes, believing that the well-off deserve to keep much of their earnings, and the poor mostly deserve only what they work for. This thinking directly contradicts the logic underlying optimal tax theory and, in turn, the efficiency-minded stance of law and economics. People think this way because they have separate policy silos for taxation and other policies, each with its own distributive views. As a result, there is imperfect fungibility across the silos. And distributive attitudes about taxation do not necessarily reflect distributive attitudes about other means or about distribution overall.

Rather, under the theory of policy silos, it is completely consistent for people to both dislike strongly redistributive taxation and wish to have more egalitarian policies toward the poor. Taxation is just one particular means for addressing inequality.

The key takeaway, if Congressional policymaking reflects the public’s views to a significant extent,159 is that taxes alone will not provide enough redistribution. People have an aversion to both redistributive taxation and inequality. Under the standard economic logic, these should be the same thing because taxes should do the work of redistribution—but that is not so with separate policy silos.

To be sure, many questions remain and urgently need more research: What are the precise contours of desert-based thinking? How important is status quo bias? How important is an appearance of “ownership” versus a sense of “earning”? How specific are these sentiments to taxation versus other owned things (e.g., property in the case of eminent domain160)? How are different types of taxation implicated? How malleable are these ideas? What do they seem to change in response to? More generally, beyond desert and beyond taxation, how else is policy siloed in political psychology, and what are the implications? Nevertheless, this Article provides a novel account that fits with what we see in the world, explaining that the public treats taxes differently as a means of redistribution.

Policymakers Should Reject the Standard Approach

The standard economic approach to law and inequality does not work and must be replaced. Economics tries to find the most efficient way to help the poor given limited resources, and it says that giving cash through taxes and transfers is typically the best way to do that. Doing so leaves a relatively clean and practical prescribed approach to law: adopt the efficient legal rule.161 Efficiency also has the appearance of “neutrality” in considering everyone’s willingness to pay equally.162

However, with policy siloing, the one-pieist method that underlies law and economics textbooks and much lawmaking163 simply ceases to be a useful heuristic in deciding any individual policy question. It works adequately with the one benevolent dictator that economists imagine in a hypothetical world that lacks politics. But once there are multiple real-world actors implementing policy in a democracy and the attitudes of voters matter in practice, the fallibility of standard economics approach comes into view. Because of resistance to redistribution in the tax silo, we cannot depend on Congress to address all distributive issues through taxes, so long as Congress is to some extent responsive to public attitudes.164 Thus, since the prescription of adopting efficient nontax laws that do nothing for inequality depends upon an unrealistic assumption about redistributiveness in taxation, that prescription does not hold.

Worse yet, efficiency is not “neutral” in the sense of providing the same legal entitlements to rich and poor.165 Since it is based on willingness to pay, and the rich tend to be willing to pay more, efficient policies tend to endow the rich with larger legal entitlements than the poor. They embed a “rich get richer” principle.166 The transportation spending example, in which more money is spent on the rich than the poor because the rich are willing to pay more to commute more quickly, exemplifies this non-neutrality. Of course, this effect matters little if taxes and transfers make up the difference. But, if there is a drag on those taxes and transfers, then the standard economics prescription for efficient legal rules leads to not just a failure to address income inequality but actually increasing income inequality. Thus, the standard approach fails on its own account of providing good prescriptions for achieving just distributional outcomes. Especially at a time of such concern over inequality, this approach is unacceptable. It has been a mistake to adopt in a democracy an uber-norm that is so unintuitive to the public.

Thus far, the Article has been a critique: The standard economic approach to law and inequality does not work on its own terms. The idea that nontax legal rules should be efficient and do nothing to address inequality ignores reality. Going forward, the Article turns prescriptive. For that, we need a new economic approach to law and inequality.

That approach begins with two stipulations. First, we need greater economic equality. This stipulation could be based on a variety of foundations: the apparent preference of the American population across the political spectrum,167 first principles in economics about the declining marginal utility of income, or first principles from one of the many other egalitarian traditions that find the current amount of inequality unacceptable.

Second, for analytical clarity, while rejecting the standard approach’s view of how policy is made in practice, the approach here retains the standard approach’s normative stance described in Part II of basing policy on the combination of efficiency and achieving a fair ultimate distribution. At the same time, the analysis going forward is meant to be an economics reference point—a useful but partial framework—rather than a fully articulated theory of justice. As such, it is meant to be broadly appealing to egalitarians. Many normative traditions168 accept some basic principles: (1) it tends be good to move resources from the rich to the poor, all else equal, (2) individuals’ choices should typically (though not always) be respected, and (3) we face resource constraints, so that there are tradeoffs in helping the poor. This Article relies on these three principles going forward. Some economists rely on only these principles.169 Other people have a different or more plural normative understanding and consider values such as legality, legitimacy, dignity, autonomy, basic rights, procedural justice, and social solidarity. Political economy concerns, such as crafting policy that is politically durable, may also be important.170 But, while the thin standard economics normative framework used here may be incomplete, it is still often an important piece of the whole and thus a useful reference point.

Beginning with these three basic principles clarifies which questions to answer, provides techniques to help answer them, and describes some of the tradeoffs in answering them. And it provides policy suggestions. It is to those questions and suggestions that the Article now turns. Far too often, the economics model is taken literally, but not seriously: prescribing its literal output even if doing so seems to not take its equality concerns seriously. The rest of the Article flips the script, taking the model seriously, but not literally, as it misses essential social realities.

A New Approach to Law and Inequality:   A Thousand
Points of Equity

Currently economics has a standard prescription: Outside of taxes, the law should be efficient, even if that means hurting the poor. The harm to the poor is justified by the assumption that taxes and transfers will redistribute enough. As this Article has shown, common psychology creates significant barriers to the adoption of these taxes and transfers in practice.171 The standard prescription is thus a recipe for spiraling inequality. We need to replace that prescription with a new one. This Article rewrites the playbook and provides a new foundation for economic policymaking in the real world.

This Part begins by illustrating how one particular institutional actor, the Department of Transportation (“DOT”), should respond to resistance to redistribution through taxation. It then develops a new approach in general, drawing lessons from both policy siloing and standard economics efficiency logic, to offer guidance on how egalitarians should promote distributive justice. This is especially relevant as the Biden Administration seeks to institutionalize equity considerations.172 Popular resistance to using the tax silo for redistribution suggests that much redistribution must come through nontax policy. Efficiency—which remains an important consideration for egalitarians—suggests that redistribution should occur at a modest level across many policies: in short, a “thousand points of equity.”

The Part continues by drawing a host of lessons from this melding of the political and the economic and describes how such a practice can be institutionalized, including through an Economic Equity Framework Act to help ensure equitable policymaking. Lastly, this Part offers concrete examples from across the law of potential “thousand points of equity” responses in areas including civil penalties, social insurance, labor law, antitrust, housing law, tort law, and specific features of tax law.

Example: Transportation Cost-Benefit Analysis

Suppose that a new Secretary of the DOT is considering how to allocate the transportation funds that Congress appropriates to it. As described in Part I, current practice is to value the time of the poor at $25 an hour and the time of the rich at $63 an hour. This system encourages spending funds on the rich instead of the poor because saving time for the rich through transportation improvements is more valuable than equivalent time savings for the poor. This regime is efficient and is thus the standard economics prescription. The government should spend the bulk of its transportation funds on the rich and then share the ensuing benefits with the poor through large cash transfers, funded by high taxes on the rich. However, that tax regime is not realistic. Income tax policy redistributes insufficiently now and likely in the future as well. What is a Secretary of Transportation attentive to this political psychology to do in spending these billions of dollars?

The DOT planners have a choice: spend inequitably and exacerbate inequality, or do something about inequality. The Appendix provides a model of how the DOT might approach addressing inequality, while remaining attentive to efficiency concerns. It then simulates what policymakers should do using some plausible assumptions about the value of redistribution, as well as the benefits and costs of transportation improvements. The estimates show that the DOT should spend roughly the same amount on the rich and the poor, treating them like they have equal wages. Indeed, as the Appendix simulation shows, the welfare impacts of a “worst of both worlds” policy—in which the poor get little cash and little transportation funding—can be very large. Regulators’ adoption of efficient policy prescriptions risks considerably reducing welfare when Congress does not also use economists’ chosen tool of taxes to redistribute. And, recall, cutting-edge research in economics suggests that the US tax code currently redistributes remarkably little.173 There is a lot of redistributive space for nontax policy to productively fill.

Importantly, the recommendation is based not on the idea that rich and poor should be treated equally in this particular policy, though that could also be important normatively (and at minimum likely makes it politically palatable). Rather, the recommendations are based on standard economics reasoning: We need to get the poor more resources somehow, and this is a relatively efficient way to do it.

Here is how the model works: There is a certain social value to redistribution based on people’s income, using a method like that of Hendren’s discussed above. For example, a reasonable estimate is that, before government policy intervenes, the social value of $1 in the hands of the poor is the equivalent of $13 in the hands of the rich. Of course, this ratio will go down after there is redistribution to the poor through taxes or other means, since the poor get a little richer.

But there is also a cost to shifting transportation funding from the rich to the poor. A quicker commute for the rich, who would then work more, contributes more to economic output than a quicker commute for the poor. There will be some projects that save a lot of time for the poor, so are worth doing in any case, and there will be some projects that save little time for the rich, so are not worth doing in any case. The question is how to handle projects that save a modest amount of time for either group. Under the “efficient” status quo, such projects would only be built for the rich, not the poor, since the rich earn more money per hour. Under the “thousand points of equity” approach, some would be built for the rich and some for the poor.

What happens in the model is that the two effects roughly cancel each other out. There is greater social value of transportation spending benefitting the poor by $1 than benefiting the rich by $1, but it also costs more in transportation spending to benefit the poor by $1 because they earn less per hour.

More equal transportation spending makes sense because it is a relatively efficient way to redistribute. In a regime that values the time of the poor at only $25 per hour, there are a lot of unfunded projects that will save considerable transportation time for the poor and thereby help them earn quite a lot of money. We should build those projects. One need not appeal to norms of procedural fairness in a specific context to arrive at the conclusion that the rich and poor should receive equal funding.

Politically attentive regulators will recognize that there is nothing “neutral” about valuing the time of the rich more than the time of the poor and thus spending more on the rich than the poor. That is efficient, but it is not “neutral.” In fact, it transparently biases allocations toward the rich. And, without the justification that taxes and transfers will address the problem of distributive equity, the rationale for purely efficient policymaking disappears. And remember, “redistribution” to the poor that just treats the rich and poor alike would not require violating a fairness norm of equal treatment. It would just mean spending equal amounts and valuing the time of the rich and the poor similarly.

Overall, what matters is not so much the specific result as the general method. Weighing the redistributive benefits against the redistributive costs suggests that the efficient status quo policy is bad. It immiserates the poor by spending less on them than the rich, while providing relatively little cash aid in compensation for the inequitable distribution of spending. Transportation planners are naïve and allocate spending as if there were large cash transfers, even though we have no such thing. Evidence suggests that access to transportation is a key factor in the ability of the poor to achieve economic mobility.174 So this spending policy considerably harms the poor. The Secretary should therefore change the current guidance that values the time of the rich at more than the time of the poor175 and instead should value their time more equally—or perhaps exactly equally.

The Approach in General

Resistance to redistribution in the tax silo means that inequality will be too high if taxes are the sole redistributive mechanism. As discussed above, desirable distributional outcomes are unlikely through taxes alone.176 So, those looking to make policy more redistributive, either out of prior normative commitments or to better match voter preferences, must turn to nontax policies. Reallocating transportation spending is one option. There are many others.

Economic insights about efficiency provide one of the starting points for a new law and economics of inequality. While egalitarians are united by their concern over distributive justice, many remain concerned with efficiency as well.177 All else equal, more efficient policies mean that there is more to go around, making everyone (at least potentially) better off. Egalitarians may be less willing than others to sacrifice equity on the altar of efficiency, but they are not indifferent to efficiency concerns. In short, tradeoffs remain important. This melding of economics and political psychology yields several lessons, which this Section discusses in turn. Of course, it is beyond the scope of one Article to describe how every actor can enact every possible policy. After all, we have no benevolent dictator. Rather, the goal is to provide some broad guidance on how to improve well-being in light of the failures the Article identifies.

A first lesson is that policymakers should adopt a “thousand points of equity” approach—redistributing modestly across many policies. Since reliance on taxes alone for redistributive aims is insufficient, there must be more than one point of equity in the system. And policymakers should look across the board to find what are likely to be some low-cost opportunities to help low-income Americans in each policy space—but not redistribute too much in any one.

Second, coordination across different points of redistribution is helpful but not essential. The more coordination, the more cost-effective redistribution will be. But what is most important is having a similar goal in mind. One way to coordinate is through an Economic Equity Framework Act, which would mandate consideration of economic equality across the range of government activities when consistent with statutory purposes.

Third, how much to redistribute depends on issues of political economy—on expectations of how policymaking will evolve in the future.

Fourth, to be most efficient, policymakers should look in unexpected places for opportunities to redistribute, but also be cautious of redistributing too much in expected places. In doing so, policymakers should sometimes consider individuals’ incomes, rather than being blind to income.

Finally, a central normative question that might seem important—whether people deserve the income that they earn—actually turns out not to matter for what to do outside of taxation. All that matters is that, in practice, taxes will not redistribute sufficiently.

Lesson One: Redistribute in a Lot of Places, But Modestly

Two important insights from economics suggest that, as a baseline, redistribution should take a “thousand points of equity” approach by redistributing a modest amount across a large number of areas, rather than a lot in a few areas.

The first insight from economics is that, for modest amounts of redistribution in any given policy, typically the redistributive benefits are large and net efficiency costs are small.178 Consider shifting transportation spending from the rich to the poor, starting from the efficient policy of maximizing total income, in which there is much more spending on the rich than the poor. As transportation planners shift money from rich to poor, they start funding the best available projects for the poor and stop funding the worst currently funded projects for the rich. So, the first shifted dollar is very similar to the efficient level of spending, producing roughly as much income for the poor as is lost for the rich and thereby resulting in miniscule net costs.

The same reasoning applies generally. Typically, the net costs of adopting modest amounts of redistribution in any given area are small and the redistributive benefits are large because of the availability of high-return ways of spending money close to the “efficient” use of resources. Thus, even for the most efficiency-minded person, the costs of modest differences from the efficient ideal to achieve greater equality are not very big. So, we can be fairly relaxed about the cost of modest deviations from efficiency and excited about the redistributive benefits.

The second key insight is that, while efficiency costs are small at first, they tend to accelerate as redistribution in a given area increases.179 Consider again shifting transportation funding from the efficient policy. As more and more funding shifts, each dollar increasingly produces less income for the poor—for example, from building bus routes in less-trafficked neighborhoods. And each shifted dollar increasingly loses more income for the rich—for example, from losing train projects in highly-trafficked neighborhoods. So, while shifting $1 million from rich to poor might result in only $10,000 in lost income overall, shifting a second $1 million might result in several times that amount in lost income because it produces little income for the poor and loses a lot for the rich.

These two insights together suggest that, as a baseline, we should adopt a “thousand points of equity” approach of redistributing modestly in many places. The first insight says that there should be at least some redistribution in lots of places because, for any given policy, there are likely some low-cost opportunities to redistribute. The second insight says that, for any given policy, that amount of redistribution should be modest, or else there can be accelerating costs and diminishing returns. For example, rather than redistributing yet more from transportation, it is better to redistribute somewhere else—ideally, many places, taking advantage of the low-cost redistributive opportunities from other policies.

That is not to say that we should redistribute everywhere. Sometimes the benefits will not justify the costs. Likewise, economic analysis will sometimes identify a large supply of projects benefitting the poor, justifying a large amount of redistribution. But, since the efficiency costs of redistributing are typically small at first and then climb for any given policy, it is better to increase redistribution a modest amount through many policies rather than increase redistribution a lot through few policies.

The thousand points of equity approach accomplishes at least two goals. First, adopting it significantly helps the poor at modest cost to the rich while avoiding policies that become so redistributive that they start doing relatively little to help the poor at high cost to the rich.180 So, if the marginal consumption of relatively rich people (say, those earning $63 per hour, in the transportation example) still counts for something, then the policy avoids doing harm to them that is not justified by very small gains to the poor. Second, even someone maximally concerned about inequality and minimally concerned about efficiency should still care about efficiency. We want to make sure that any given policy does not do too much harm to the economy, thereby limiting the resources to redistribute through other policies. With the thousand points of equity strategy, the poor can get more because the rich lose less from any given policy. Thus, the rich have more to redistribute through other policies. Notwithstanding the presence of policy siloing as a matter of how ordinary people think, there is still only one economy, but lots of policies that we can use to redistribute.

Lesson Two: Coordination Is Helpful, But Not

Coordination across the various parties who could implement a thousand points of equity approach—agencies and legislatures at the federal, state, and local levels—would be helpful, though not essential. Ideally, as many parties as possible would have in mind a common tradeoff between a dollar (or dollar’s worth of goods or services) in the hands of the poor and the rich—for example, a ratio of five to one for those at the tenth versus ninetieth percentiles of income. This is, of course, a normative judgment. With this ratio, if economic analysis shows that a policy would benefit the poor by $1 at the cost of $5 to the rich (an efficiency cost of $4), it would still be worth doing because of the distributional benefits. Accordingly, the various parties could engage in the same tradeoff of distributional benefits with efficiency costs to ensure that the most efficient means of redistribution are adopted, conducting analysis like this Article does for transportation spending. The goal of this Article is not to fully articulate how to best achieve coordination. But it can begin to sketch out possible strategies.

Such an explicit full coordination is not practical across all parties. But consider how federal agencies might coordinate as one example. One way to institutionalize this concern with equality across many domains would be to enact an “Economic Equity Framework Act,” requiring administrative agencies to interpret statutes to achieve economic equity, in ways consistent with statutory purposes.181 The more precise a tradeoff between equity and efficiency that is stated, the better. But even a general statement about the value of equity, without undue harm to the economy, could help get the various parties engaged in considering efficient opportunities to achieve greater equality.

A framework statute could be supplemented by guidance from the White House Office of Management and Budget that gives a specific value of redistribution throughout the federal bureaucracy based on an executive order from the President182—or at minimum a greater mandate to consider economic equality than currently exists.183 This consideration of distributional impacts could operate much like current federal guidance on “the social cost of carbon,” which gives a specific dollar cost for each ton of carbon emitted to use in measuring regulations’ climate impacts.184 This guidance could also be a substitute for a framework statute. However, guidance alone would not be ideal because courts, finding a lack of statutory authority, or subsequent presidential administrations, could potentially overturn it.

Legislatively, the Congressional Budget Office could formalize a process of measuring distributional and efficiency impacts for proposed legislation, which would help Congresspeople and their staffers make appropriate tradeoffs. Subnational governments could make parallel changes.

All this said, the “thousand points of equity” approach does not require central planning. Of course, the more parties involved and acting with common purpose, the less costly redistribution will be, by allowing policymakers to take advantage of the most cost-effective opportunities. Still, no explicit coordination per se is necessary. Parties can act independently.

The actors can also respond dynamically. The more redistribution there is elsewhere, the less it is worth any actor doing. For example, a particularly sophisticated Office of Management and Budget could potentially have triggers that ratchet down redistributiveness if programs elsewhere redistribute more or economic circumstances change, much like extended unemployment benefits are triggered when the unemployment rate increases.185 However, a crucial feature of having a thousand points of equity is that each actor is typically quite small relative to everyone else. So, how one party acts will typically have little impact on how others should act. And again, each party can act on its own. 

As with any policymaking, one would still need to coordinate the substance of policy—to ensure, for example, that one part of the DOT does not fund a bus line while another funds a repetitive light rail line for the same place. But that is already true. So, little fundamentally changes along the substance of coordination if policymakers try to be more equitable.

Of course, these changes would need to operate against a backdrop of other values beyond the scope of this Article. For example, explicit redistributionist goals may be problematic on procedural justice grounds. And, nothing here is to say that a thousand points of equity would not be messy. It would be! The proposal is for lots of actors to help low-income individuals get a leg up, rather than having the clearer but perverse goal of efficiency that often actually harms the poor. Undoubtedly, different actors will respond differently. But democracy, which has no benevolent dictator, is messy. This is our reality. And policymaking should be sensitive to this reality—making policy based on the world as it is, not on the world as some economists imagine it.

Lesson Three: How Much Redistribution Depends on
Political Economy

In pointing to the political nature of policy adoption surrounding redistribution, the Article shows the importance of political factors affecting the amount of desirable redistribution. In particular, there is a risk that less efficient redistribution today discourages more efficient redistribution later on.186 For example, Congress may simply see more redistribution elsewhere and, in response, reduce redistribution through taxation.187 To the extent that is true, new policies should redistribute less. Partly because of policy siloing, there is reason to suspect that this factor is not large. For example, rigorous empirical work shows that more redistribution through schools has zero effect on redistribution through taxation.188 Nevertheless, this is an empirical parameter in need of more measurement.

Lesson Four: Look in Unexpected Places for Equity

To redistribute most efficiently, policymakers should look in unexpected places for opportunities to redistribute. They should also be cautious about redistributing too much in any one place, even prominent ones. Importantly, the thousand points of equity approach does not call for simply treating individuals identically regardless of income. Rather, the goal is to redistribute modestly relative to an efficient policy. Sometimes the approach will suggest blindness to income, as in the transportation example; other times, not. Two examples help illustrate.

Consider the example of fines. The classic efficiency argument is that fines should not be based on income.189 So, if a rich person speeds, she gets a $200 ticket. And, if a poor person speeds, she gets the same $200 ticket, even though it causes greater hardship to the poor person.190 The reason for the identical fine is that the harm from speeding—in efficiency terms—is the same regardless of the income of the speeder. The rich person might choose to speed and pay the $200 fine, but that speeding is presumably worth at least $200 to the rich person. Assuming that the fine is set efficiently (that is, based on the harmed parties’ willingness to pay to avoid the speeding), the expected harm from her speeding is only $200. So it is efficient for the rich person to speed and pay the fine, since she values speeding more than the harm that it causes.

But the desirability of efficient fines depends upon policymakers treating distributional outcomes through fines and taxes fungibly. If instead there is a drag on tax-based redistribution, then charging the rich more than the poor for speeding or other civil infractions is good policy, even on standard economics grounds. Inefficiency results from income-dependent fines: the rich will drive “too” cautiously as a result. But, essentially since the poor feel the pain of a given amount of fine more acutely than the rich, it makes sense to have higher fines for the rich.

At the same time, some egalitarians may find that everyone should have access to the best-available healthcare without paying anything out-of-pocket—another way of treating everyone the same regardless of income. However, a thousand points of equity approach would caution against that, as recipients of much of the healthcare provided, especially to lower-income individuals, may not value the healthcare at nearly the amount that it costs to provide and would prefer to receive those resources in other forms where redistribution is not as high.191 In short, the distributive benefits may not justify the efficiency costs because recipients do not actually value the services very highly.

Of course, lawmakers could adopt policies requiring equal treatment regardless of income in both cases—identical fines for everyone and free provision of the best-available healthcare for everyone. Indeed, that might make sense, depending upon one’s values.

But there is a cost to such income blindness. Though voters may not often see things that way, it is fundamentally true that as a society we face tradeoffs. Notwithstanding the falseness of the idea that policymaking as a whole operates as if policies are fungible, in fact we have just one set of resources. So, we can only redistribute so much. More redistribution in one way places limitations on redistribution in other ways.

An income blindness approach has efficiency costs compared to the thousand points of equity approach because in some cases (such as fines) it redistributes “too little” and in other cases (such as free best-available healthcare) it redistributes “too much.” Averaging out legal rules with insufficient redistribution with legal rules with excessive redistribution would create costs for two reasons. First, doing so might redistribute the wrong amount. Second, the approach would miss efficient opportunities to redistribute, which could make the poor worse off by providing them with fewer resources.

In short, in providing a foundation for policymaking, the thousand points of equity approach can suggest good places to look for redistributive opportunities, channel intuitions (for example, toward income-dependent fines and away from free best-available healthcare), and provide guidance on how much to redistribute in any given case. With data on the efficiency cost of income-dependent fines and a stated value of redistribution, it can suggest how progressive the fine structure should be. Of course, other values like retribution, which are also promoted by such fines, could also matter for policy design. But the goal here is to point out that—on distributive and efficiency grounds alone—it is a good policy to consider, and here is an approach to designing it.

Lesson Five: Whether People “Deserve” Their Income
Does Not Impact Recommendations

The thousand points of equity approach is a better baseline than the efficient one, regardless of one’s views about the normative status of desert to keep pretax income. Siloed tax policy attitudes about desert raise a host of deeper normative questions: Does pretax income have moral weight? Does desert inhere in effort, in inborn talent, in talent that one works for? Does taking individuals seriously as individuals mean that the government should not tax based on genetic predisposition to earn money? This Article cannot answer those questions. And it does not need to in order to advocate for the thousand points of equity approach. Of course, the normative status matters for the kind of income tax policy that one would ideally want to adopt. If people actually deserve to keep the income they earn, then more redistributionist taxation is less normatively desirable. This kind of debate about the relation between moral intuitions and ideal policy is a common one—for example, in torts.192

But the question here is different: What is the impact of views about income tax policy on the other policies that we should adopt? For that, the thousand points of equity approach is equally appealing whether or not people deserve their pretax income. The reason is simply that, so long as redistribution is insufficient to address demands for equality, we will need to look elsewhere to achieve equality goals. This is true regardless of whether or not it is good to redistribute a lot through the particular means of cash taxes and transfers. For those who are formally inclined, the Appendix presents a model showing this result.

Here is the intuition: On one hand, if people are durably mistaken in their belief that pretax income generates desert, then policymakers will need to look elsewhere to achieve distributional goals. In this case, the thousand points of equity approach is “second best.” There is a better approach, but it’s just not politically feasible.193 On the other hand, if people really do deserve their pretax income, the conclusion is the same. In this case, the thousand points of equity approach is “first best.” The approach results not from a political constraint but rather because it is the right thing to do for policymakers to not redistribute as much through taxation. But that doesn’t change the fact that greater overall equality is still a goal, notwithstanding the particular distaste for using income taxes. This distaste just means that policymakers should look elsewhere to achieve equality.

In both cases, efficient but inegalitarian nontax policies are typically unjustified. Instead, nontax policies should typically redistribute (and do so to the same degree) because the same amount of redistribution is needed in either case. The Article is agnostic on whether people deserve the income that they earn, and indeed a powerful point of the Article is that the conclusion about nontax policies is the same either way.

Other Legal Applications

The thousand points of equity approach opens up wide frontiers in the law for addressing inequality. The standard approach in many policy areas is to do nothing at all to address inequality.194 Under the standard approach, adopting something other than the efficient but inegalitarian law, while perhaps intuitive, is soft-headed. According to its proponents, the better approach is to maximize the size of the pie and then redistribute through cash taxes and transfers. This Article flips the script: Laws that address inequality are often the right ones precisely because they are hard-headed—they reflect the reality of policy siloing.

This Section offers egalitarian legal changes that are unwise under the standard approach but are supported by the thousand points of equity approach, as they modestly redistribute at low efficiency costs. Importantly, the recommendation is not to sneakily adopt low-salience policies that violate the public’s sense of what is fair, but rather to be open to redistribution where those other norms allow or even encourage redistribution. To reiterate, a key feature of policy siloing is that people feel differently about redistribution in different domains, and just because people resist redistribution through taxation does not mean that they oppose redistribution elsewhere.195 And policy siloing certainly does not mean that people want efficient policies. Of course, one Article cannot incorporate every possible value, public intuition, and political consideration. But it can point to some possible fruitful areas for policy development, which include (among many others across the law) regulation and torts, social expenditures, and tax policy itself.

Regulation and Torts

Regulatory cost-benefit analysis: Beyond transportation spending, current federal regulatory cost-benefit analysis is typically based on efficiency to the detriment of distributional impacts.196 This regime is supported by economics-oriented scholars, on the grounds that taxes and transfers will undo the distributional consequences of regulations.197 The theory of policy silos explains why this is a bad premise and provides guidance on the extent to which distribution should be considered. The more we fail to redistribute through taxes, the more we should do so through regulations.198

Pro-poor regulation in areas like collective bargaining or antitrust: The cases for a variety of regulatory changes that benefit the poor are strengthened by resistance to redistribution through taxation. For example, some propose strengthening labor law to improve the ability of workers to collectively bargain.199 Such changes could harm efficiency and be subject to the critique that workers should work without strong unions and then money can be redistributed after wages are earned. One could make similar statements about considering distributional impacts in antitrust law200 or a minimum wage.201 The case for such regulations is substantially strengthened by this Article. Modest legal adjustments—say, a modest minimum wage or antitrust rules that stop mergers in the cases with the worst effects on low-income workers and least benefits for others—could yield significant benefits to lower-income individuals at only a small cost, while also being intuitively appealing to the public.

Torts and damages: A widespread movement has cut back on tort awards for pain and suffering,202 even in the face of evidence suggesting that juries used to provide larger damages to poorer groups.203 Statutes could return to juries the latitude to conduct such redistribution. Additionally, “economic damages” in torts are typically based on lost wages, so that a higher-paid person receives larger damages than a lower-paid person.204 This policy is efficient because people are willing to pay for their wages. A similar mechanism was at play in the September 11th Victim Compensation Fund, in which higher-paid victims received higher compensation.205 Tort rules could be modified to reduce the reliance on income to reduce these disparities.

Social Expenditures

Social insurance like parental leave and childcare: Standard logic suggests that low-income parents would likely prefer to receive cash instead of either subsidies or directly provided childcare. Childcare is expensive, and low-income people need cash.206 Of course, there may be important standard efficiency reasons for childcare provision, such as increasing children’s lifetime earning capacity, for which parents may be unwilling or unable to pay. But, even apparently “inefficient” childcare is good if (i) redistributive taxation is limited and (ii) the willingness to pay of the poor is close enough to the cost to justify the expense. For example, if the poor would be willing to pay $5,000 for childcare, but it costs $8,000 to provide, provision very well may make sense because the social value of directing resources to the poor is high. Similar reasoning can be applied to spending on education, paid parental leave, nutrition, and a panoply of other social insurance programs.

Necessities accounts: Some direct transfers can redistribute cash without appearing to do so. Consider, for example, a “necessities account” that can only be used on necessities like food, utilities, commuting, housing, clothing, education, and healthcare. It would be functionally similar to a cash transfer without looking that way because the poor spend most of their money on those things anyway.207 And, as noted earlier, a necessities account is considerably more popular than an unrestricted cash program.208 The account could be phased out at higher incomes, and it would effectively function like a cash transfer to the poor.

Recasting in-kind support like housing: At least from the standard perspective of economics, widespread government support for in-kind transfers like housing is surprising.209 Why not just provide cash? While the right to housing or the desire to ensure that children live in good conditions may provide compelling reasons, another reason is that in-kind transfers represent a politically popular alternative to cash.210 Everyone needs housing, so providing at least some housing support is not likely to encourage recipients to spend hugely more on housing relative to other things that they need. If an important reason for providing housing support is not really about funding housing in particular but is instead about finding a popular alternative to cash, then certain design elements come into focus: Support should be flexible to use and focused on helping more people rather than a smaller benefit to only a share of the poor, as is the case now.

Lowering social insurance taxes that employers pay on low-income workers: Given the aversion to cash transfers to low-income workers, an attractive alternative with a similar effect may be lowering social insurance taxes that employers pay upon hiring low-income workers.211 We are accustomed to thinking of employees as earning a salary and then paying taxes based on that salary. In fact, employers also pay taxes based on their employees’ salaries, through payroll taxes that nominally fund Social Security, Medicare,212 and unemployment insurance.213 These taxes have likely played a significant role in reducing workers’ wages.214 Reducing these taxes on low-income workers could drive up demand for these workers, increasing their wages without running afoul of desert norms—since it would just be letting employers keep more of the money they earn.215

Tax Policy to Raise Revenue

While some changes (e.g., reallocating transportation funds or changing some legal rules) would not cost the government money, others would require additional revenue.216 There are many options for doing so even given desert-based income tax norms:

Raising taxes that employers pay on high-income workers: Congress could raise taxes that employers pay on their high-income workers. Such a tax would presumably lower the pretax wages paid to these employees because employers would have to foot a larger tax bill, reducing demand for such workers.217 Indeed, the tax code already has some features like this. For high-income executives, the tax code limits the amount of compensation that publicly-traded employers can count as business expenses—notably one of the few areas in which taxes on the rich increased in the 2017 tax bill.218 Of course, there would be implementation difficulties, such as the incentive for a high-income employee to split his job at one company into two lower-paid jobs at two companies to avoid the tax. But, if voters read such taxes largely as being on businesses, rather than on the earners themselves, who may seem more deserving, then it may be tolerable to impose such taxes on employers in a way that raises revenue that is targeted at the rich.219

Taxing capital at higher rates: Recall that under standard economic theory, even though the rich tend to get more of their income from capital than the poor do, taxes on capital should be zero or low, since they are viewed as less efficient than taxes on labor.220 Federal income tax rates on capital are, in fact, typically considerably lower than those on labor.221 But, if labor income taxes stay relatively low, Congress could raise capital income taxes on the rich—perhaps to the point where capital income no longer had a lower, preferential rate.222 Indeed, there are many options for taxing capital, which could raise trillions of dollars.223 Taxes could be placed on individuals who own parts of businesses or on the businesses themselves, for which evidence suggests strong popular support.224 Indeed, polling evidence suggests that most support taxing capital income at the same rate as labor income.225 This fits with what one would predict with desert-based taxation, since, while saving may be virtuous, it probably does not generate more desert than exerting work effort.226

Higher consumption taxes: Alternatively, the United States could do what almost every other rich country has done and adopt a national “value-added tax,” which is similar to a tax on consumption.227 Consumption taxes are regressive, since poorer people consume a larger share of their income, but not nearly as regressive as social spending is progressive, so a value-added tax to fund more social spending would go far to reduce inequality.228 And crucially, taxpayers may seem more deserving of their income than their consumption, which has a more attenuated relationship with work. Indeed, this combination of only modestly progressive taxes and very progressive spending is exactly what the Nordic countries have employed, while becoming some of the most progressive—and also rich—countries in the world.229

Higher state and local taxes: Redistribution by state and local governments has typically been criticized by scholars as distortive of location decisions by households and businesses. In the face of insufficient redistribution through a national-level income tax, local governments could fruitfully play a larger role in financing social spending.230

The implications of the thousand points of equity approach are sweeping. Across the law, policymakers should consider opportunities to make policies more equitable, even if it means being less than perfectly efficient.

Addressing Critiques

This Part addresses several potential critiques about both the descriptive and prescriptive contributions of the Article. While these critiques can lead to important qualifications, and working out a new law and economics of inequality is the work of a generation rather than one Article, the critiques leave in place the basic prescription of widespread redistribution to promote greater equality.

Alternative Explanations for Lacking Redistribution
Through Taxation

One might be concerned that public resistance is not the mechanism explaining the observed lack of redistribution through taxation. Nothing here implies that this is the only mechanism, just that it is important. Nevertheless, recalling that the central point here is that there is a greater public aversion to redistributing through taxes than other means, it is worth explaining why other potential explanations are incomplete descriptively and as bases for prescriptions.

One alternative explanation is capture of the political system by powerful, moneyed interests.231 On that account, we do not have enough redistribution through taxes because powerful, moneyed interests stop it. This could easily be an important explanation operating alongside the policy silo explanation. But it is probably not the full story. Crucially, it does not explain why we see so much redistribution outside of taxation. There is overwhelming evidence that people care about more than their own self-interest in forming policy preferences.232 They also care about fairness. Partly as a result of these fairness preferences, outside of taxes on earned income, we do see egalitarian policymaking and rhetoric.233 As described earlier, people care a great deal about access to necessities. Even Republican politicians talk about universal access to healthcare, for example.234 And we see concerns with inequality in a host of other places as well, such as when regulatory cost-benefit analyses value the lives of rich and poor equally. While acknowledging that we need to know more, it seems like there is particular resistance to redistributing with cash taxation based on earned income.

Another potential explanation is that people just do not care much about economic inequality. But, polls consistently show the opposite.235 So does our politicians’ rhetoric, as both parties decry inequality and a rigged system.236 And so do all the egalitarian attitudes that were just listed, such as the right to healthcare declared across party lines.237 Rather, as the theory of policy silos says is the case, people have different views on overall distribution than on redistribution through taxes specifically. In particular, Gallup polls have long found that a considerably larger share of the public believes that the distribution of income should be fairer than believes that heavier taxes on the rich should be used to achieve that greater equality.238 Resistance to redistribution through taxation specifically is not the same thing as resistance to greater economic equality, which receives strong support.

A third possibility is an institutional failure in which, if institution A (e.g., the courts) makes a policy change with distributional consequences, institution B (e.g., Congress) does not compensate for those consequences, out of inertia or otherwise. Work in law and economics commonly discusses this critique about a potential failure of taxes and transfers to compensate for distributional changes.239 However, while this is likely an important phenomenon, it does not explain the depth of taxes’ failure to redistribute. Suppose that compensatory tax changes did follow the adoption of every nontax policy. Even putting aside that some of those tax changes would help the rich, redistribution could still be massively insufficient because of a background failure to redistribute enough through taxes. Considering policy responses one-by-one does not ask the most important question if society is not distributionally just to begin with.

The mechanism for inequality matters, since it affects the prescription. If a deal is on the table to make a nontax policy more efficient by hurting the poor, but then compensate by taxes and transfers, making everyone better off, policymakers should typically take it. But there is another step if there is a background failure to redistribute because of anti-tax attitudes: Additionally, when they are able, policymakers should typically make that nontax policy more generous to the poor again, at least if they can do so without significantly disrupting future deals to make everyone better off. So, the policy recommendation remains: Redistribute to the poor through nontax policy.

Why Might Policy Silos Lead to Too Much Inequality?

The theory of policy silos raises something of a puzzle. If public attitudes in the tax silo hinder redistribution through taxes, but people still want more redistribution, why do we not see enough economic equality as judged by voters,240 especially given pro-redistribution attitudes about categories like necessities in other silos?

We cannot know, but we can speculate. The first reason is institutional. Voters’ opposition to redistribution through taxation may play a larger role in setting tax policy than voters’ support of redistribution through nontax means does in setting nontax policy. The reason is the unusually large role of Congress in setting tax policy,241 with the tax code specifying exact tax rates for exact incomes,242 which are typically set in heated, public political battles.243 In other policy areas, experts in agencies and judges on courts, with potentially less responsiveness to voters, may hold more sway.244 For example, with transportation spending, Congress appropriates money to the executive branch, which often decides how to allocate the funds with little guidance.245 Indeed, for many regulatory and spending programs, Congress delegates to agencies a major role in distributional decisions, whether because of concerns about expertise, limited capacity, reducing logrolling, or otherwise.246 Of course, Congress and legislatures are quite prescriptive in some areas, such as provision of health insurance and housing vouchers. But in many other nontax areas—for example, antitrust law,247 tort law,248 and environmental law249—agencies and courts hold great sway. If those nontax experts focus on efficiency, perhaps because of their economics-oriented training,250 then the combination of efficient nontax policy and insufficiently redistributive taxes (disproportionately reflecting lay resistance to redistributive taxation) would yield insufficient redistribution overall.

A second explanation is that there is nothing requiring coherence among lay policy attitudes. Nothing requires similarly strong beliefs across different domains. People may have strong views about the one big, high-salience issue that is supposed to be used for redistribution—taxes. In contrast, people may simply have less strong views on average about the myriad and often low-profile nontax issues, as it may be difficult to have strong views on so many issues. So, even if voters feel strongly about inequality, if courts and agencies do not act to instantiate those values in the areas over which they have authority, overall economic outcomes may end up unequal because of tax policy to which voters are attentive. 

What if Public Attitudes Change?

Commonplace attitudes change—sometimes dramatically. Nothing in this Article suggests that resistance to redistributive taxation is innate. Suppose that public attitudes change, allowing taxes to become more redistributive. As Subsection IV.B.3 touched on, if there is inertia in policymaking, then policy could end up in a combination of excessive and inefficient “second best”251 redistribution.252 In this case, current nontax policy should redistribute less.

However, there are several reasons to redistribute more today, pushing in favor of even second-best redistribution. First, policy is not completely inertial and, if taxes become more redistributive, policymakers could just change their nontax policies to become more efficient and less redistributive, as noted in Part IV. Second, given standard discount rates, redistribution today matters a lot relative to hypothetical future policy. Even at an eight percent annual discount rate, $1 in 10 years is worth only $0.45 now.253 So, the costs of waiting for views to change to allow for the most efficient means of redistribution are high. Third, we do not know in which direction policy views will head. Policy views could move toward or away from one-pieism. For example, over the last few decades, policy has in many ways moved away from the kind of cash provision suggested by economics and toward work requirements.254 Nevertheless, the risks of inertia in inefficient redistribution do remain a consideration against second-best redistribution.

A related concern is that attitudes can change through education and that such education is a good alternative to adopting such second-best policies. That is, if one is convinced that desert-based tax views result from misunderstanding rather than valid normative judgments, then the first response should be to educate people. However, without denying that education can have an impact,255 there is strong reason to think attitudes against redistributing through taxes are durable. Indeed, while some of the specific implications of optimal tax theory have only more recently come into focus, arguments on similar grounds for cash transfers date at least to Milton Friedman in the 1960s,256 and optimal tax theory itself dates to the 1970s. The project of economics—especially law and economics—over the past several decades has been one of educating about efficiency and tradeoffs.257 Over all this time, tax institutions have not appeared to change much to incorporate these old ideas, consistent with their cutting against significant aspects of social psychology.258 Nevertheless, the more educable people seem about policies that we are certain are the correct ones, the less it makes sense to adopt alternatives to those correct policies today.

The Contracting-Around Problem

For some policies, like transportation spending, the affected parties do not bear the costs of policies. If poor people receive more federal funding for transportation, they do not have to pay more. For other types of policies, where there is a contractual relationship, intended beneficiaries may effectively bear the costs, making the intended redistribution less valuable.

Consider two such examples. First, consider the “implied warranty of habitability,” which requires landlords to maintain rentals up to certain standards. It may on its face seem to benefit the poor.259 But, since such requirements may lead landlords to raise rent, low-income lessees can end up worse-off. Second, some regulations impact the prices that parties pay or the wages that they receive,260 such as workplace regulations that could reduce demand for workers and lower their wages.

Complex empirical questions are involved here. But, to the extent that higher prices or lower wages harm intended beneficiaries, these areas are less promising for finding fruitful redistribution. The stronger cases are where the government provides the funding or sets background legal rules, like those for torts, when parties are not themselves in a contractual relationship that allows price shifts to mitigate distributive gains.261

Uncertainty About the Right Distribution of Resources

One may also object that there is uncertainty about the right distribution of resources, which makes attempts to redistribute likely to end up at the wrong level. It is the case that, for nontax legal rules, the standard paradigm neatly elides the question of the right distribution by making taxes and transfers the sole domain of distributional concerns. However, Part III of the Article shows that this bifurcation does not work because of policy siloing. And uncertainty about the right distribution does not imply a default to the standard economic prescription of efficient nontax policymaking that is inattentive to equity, as that leads to too much inequality.

Indeed, the thousand points of equity approach suggests how to address such uncertainty. Even if we are unsure whether taxes redistribute too little, we should still redistribute through nontax means. The reason for still redistributing under these circumstances is the same as why it is good to redistribute a modest amount across many policies: The costs of deviating from the ideal policy increase as one gets further from the ideal. So, if we are unsure which of two policies is ideal, adopting a policy close to the middle of the two possibilities is prudent, so that there are not large deviations from either. Policymakers could then tailor redistribution by comparing the risks of doing too much and too little, perhaps through reasoning resembling “error cost” analysis in antitrust, which compares the costs of erroneously allowing and erroneously preventing a merger.262

Part of the issue here is that economics not only is unable to answer what the right distribution is but also has never even given a clear answer on how to approach the question. As Part II described, this answer could come from external first principles or from voters’ preferences.263 But first principles may not give clear and widely-accepted answers. And, on top of the difficulty in general of knowing voters’ preferences, policy siloing raises the specter that voters may have preferences about overall redistribution and preferences about distribution on a policy-by-policy basis.264 So, policy siloing makes the task for those who wish to rely on voters’ preferences all the harder.

This Article stipulates that we need more redistribution. It then describes how to proceed given a certain desire for redistribution. But even if one disagrees that there is too much inequality, the Article’s descriptive contribution still stands: Law and economics must be fundamentally rethought. There is little reason to expect that taxation alone will achieve the right distributional outcome, whatever that is.

Legitimacy of Redistribution Outside of Congress

One might also argue that “redistributing” through institutions other than Congress is illegitimate, as that would mean doing something that Congress or the public itself does not want.265 It is indeed the case that policy siloing makes questions of legality and legitimacy—which seem less urgent under the neutral-seeming goal of efficiency—come roaring back. Questions that seemed answered no longer are.

Across the law, there is no normatively neutral place from which to set distributional goals. All policies have distributional consequences. Making some choice is unavoidable. And it is not at all clear that either Congress or the public wishes for no “redistribution” outside of Congress. Recall that “redistribution” is a term of art here, and “efficiency” is hardly neutral. Economists measure “redistribution” relative to an efficient baseline.266 In the example of the DOT’s allocation of spending, the efficient rule is to spend more on the rich. But Congress has not dictated to the DOT how to allocate the funds between the rich and the poor.267 And it is not at all clear that voters or Congresspeople want to value the time of the rich more than the time of the poor or to spend more on the rich. This is especially so because efficient cost-benefit analysis is typically not Congressionally mandated.268 And why would it be? As the theory of policy silos explains, people have different distributive views on different issues. There is little evidence that the public believes in the prescription of heavily redistributionist taxation but efficient nontax policy.

But the transportation spending example is a relatively easy one. Redistribution there just means spending similarly on the rich and the poor. What about harder cases, in which regulators spend more on the poor than the rich or judges affirm principles of economic equality (even at a high level of generality) in addition to the principles of efficiency that have helped lead to the inegalitarian status quo? At this point, while there are longstanding academic debates,269 we cannot say what the public would find legitimate, since little research has been done on what people actually think about administrative “redistribution” or judicial affirmations of economic equality.270 By developing the theory of silos, the Article opens up wide frontiers in understanding what legitimate redistribution is in particular contexts, now that the standard justification for efficient decision-making no longer holds.


The standard policy approach among the economic policy elite prescribes addressing inequality only through taxes and transfers—and for the rest of the law to be efficient but inegalitarian. This approach depends upon empirical assumptions that are deeply at odds with social reality. The result has been skyrocketing inequality.

It is not realistic for one type of policy—taxation—to take on the burden of all needed redistribution. Rather, people silo their policy attitudes, and—while they support redistribution in areas like necessities—the public’s political psychology hinders redistribution through taxation. As a result, in the United States at present, taxes are unlikely to redistribute enough. Therefore, the economics heuristic that nontax policy should not redistribute dooms us to a harmful amount of inequality and should be rejected.

We need new foundations for law and economics that combine insights from economics about efficiency with data-driven insights about real-world policymaking, especially considering the psychology of ordinary citizens. The thousand points of equity approach, introduced here, points the way forward for addressing alarm about inequality. By modestly redistributing across many policies, we can have the most redistributive gain at the least cost, given limited resources and inevitable tradeoffs. The standard economic policy prescription puts a distributive straitjacket around all policy outside of the income tax. It’s time to take the straitjacket off.

For the past few decades, standard economic expertise has not yielded robust economic growth,271 greater economic equality,272 or greater trust in government policy.273 Populist revolts around the world have put economic policy up for grabs.274 It is time to try something different. To help do so, this Article opens up a host of questions that must urgently be answered—about the political psychology of policy siloing and about merging those concerns with standard efficiency concerns—to most effectively promote the public good, especially for the neediest among us.

The future of economic policymaking and analysis is one of wide frontiers. It will include skepticism about any one totalizing approach and questions about the legitimacy of using various institutions to achieve distributive justice. But that’s democracy. And the policies that our democracy deserves are not those generated by a blinkered benevolent dictator, but rather those that embrace the richness and messiness inherent in self-government by the people.      275

Appendix – Model of Regulatory Cost-Benefit Analysis
for Transportation Spending

Consider a new presidential Administration conducting regulatory cost-benefit analysis for transportation spending to allocate funding between the rich and the poor. This Appendix model shows how the Administration can incorporate concern for both efficiency and distribution, given that Congress redistributes less through taxation because of a distaste for redistributing through the particular means of cash taxes and transfers. The model uses a standard welfare framework, apart from this distaste for redistribution through taxation. In other words, taxes are treated as one among many possible means of redistribution, and people may have preferences over these means. This is a way of reconciling (i) a distaste for using taxes and transfers to redistribute because of desert-based beliefs with (ii) widespread concern about economic inequality. At the same time, the model allows for tradeoffs: in particular, the declining marginal utility of income driving redistribution to the poor versus the relatively low and diminishing returns to investment in infrastructure for the poor.

Three main insights come from the model. First, the model shows a concrete framework for how policymakers can think conceptually and practically about distributive questions given the reality of political constraints. Second, under the theory of policy silos, the DOT’s current policy is mistaken. Desert-based tax views lower redistribution through that means, but people still care about inequality. The DOT errs in not equalizing its transportation spending to help reduce inequality. And the welfare losses of erroneously spending only the “efficient” amount on the poor can be large. Third, one can remain agnostic on whether considering desert in taxation is normatively correct. The transportation planner’s right choice is unaffected.


  1. [1]. Emmanuel Saez & Gabriel Zucman, The Rise of Income and Wealth Inequality in America: Evidence from Distributional Microeconomic Accounts, 34 J. Econ. Persps. 3, 13 (2020) (showing greatly increased income inequality today versus the mid-20th century).

  2. [2]. See, e.g., Pew Rsch. Ctr., The Generational Gap in American Politics: Wide and Growing Divides in Views of Racial Discrimination 18 (2018), [] (noting that 62 percent of survey respondents felt that the U.S. economic system unfairly favors powerful interests, and 82 percent felt that income inequality in the United States was either a very big or moderately big problem).

  3. [3]. Republicans have increasingly expressed concern about rising inequality, even before President Trump’s populist rhetoric during the 2016 campaign. For example, Mitt Romney, Jeb Bush, and Rand Paul all complained about the continuing rise in inequality during the Obama presidency. See, e.g., Catherine Rampell, Republicans Have Started to Care About Income Inequality, Wash. Post (Jan. 22, 2015), [].

  4. [4]. See generally Raj Chetty, Nathaniel Hendren, Patrick Kline, Emmanuel Saez & Nicholas Turner, Is the United States Still a Land of Opportunity? Recent Trends in Intergenerational Mobility, 104 Am. Econ. Rev. 141 (2014) (discussing the decline of intergenerational mobility in the US).

  5. [5]. U.S. Dep’t of Transp., Memorandum from Vinn White on Revised Departmental Guidance on Evaluation of Travel Time in Economic Analysis 2 (Sept. 27, 2016), https:// [] [hereinafter Dep’t of Transp. Memo] (providing recommended hourly values of time travel savings).

  6. [6]. Gillian B. White, Stranded: How America’s Failing Public Transportation Increases Inequality, Atl. (May 16, 2015), [] (describing how living in areas with poor public transit limits access to jobs for those with lower incomes).

  7. [7]. Mikayla Bouchard, Transportation Emerges as Crucial to Escaping Poverty, N.Y. Times (May 7, 2015), [] (describing work by Raj Chetty and co-authors in which “commuting time has emerged as the single strongest factor in the odds of escaping poverty”).

  8. [8]. See generally Louis Kaplow & Steven Shavell, Fairness Versus Welfare (2006) (describing the approach); see also infra note 36 and accompanying text (discussing “Kaldor-Hicks efficiency”).

  9. [9]. Louis Kaplow & Steven Shavell, Why the Legal System Is Less Efficient Than the Income Tax in Redistributing Income, 23 J. Legal Stud. 667, 667 (1994); A. Mitchell Polinsky, An Introduction to Law and Economics 127 (2d ed. 1989); Robert Cooter & Thomas Ulen, Law & Economics 7–8 (6th ed. 2012).

  10. [10]. Kaplow & Shavell, supra note 9, at 675.

  11. [11]. N. Gregory Mankiw, Principles of Economics 145 (7th ed. 2015).

  12. [12]. See, for example, the large number of economists at think tanks. See, e.g., Experts, Brookings Inst., []; Our Scholars, Am. Enter. Inst.,; Experts, Peterson Inst. Int’l Econ., [].

  13. [13]. See, e.g., Council of Economic Advisers, Staff Economist, White House, https://www.white [].

  14. [14]. See Exec. Order No. 12,866, 3 C.F.R. § 638 (1993) (requiring cost-benefit analysis based on efficiency in federal agencies). See generally Tyler Cowen, Where in the Federal Government Do the Economists Work?, Marginal Revolution (Jan. 17, 2011, 8:11 AM), https://www.marginal [] (describing the outsized influence of economists on government and explaining there were 4,130 individuals with “economist” in their title in the federal bureaucracy in 2008, excluding the Federal Reserve); Justin Fox, How Economics PhDs Took Over the Federal Reserve, Harv. Bus. Rev. (Feb. 3, 2014), [] (describing how economists took over key Federal Reserve leadership positions).

  15. [15]. Zachary Liscow, Is Efficiency Biased?, 85 U. Chi. L. Rev. 1649, 1651 (2018); see also Cass R. Sunstein, The Cost-Benefit Revolution 6–21 (2018) (discussing the “Triumph of the Technocrats” across the Reagan, Clinton, and Obama administrations); William M. Landes & Richard A. Posner, The Influence of Economics on Law: A Quantitative Study, 36 J.L. & Econ. 385, 424 (1993) (“[T]he growth in the influence of economics on law exceeded that of any other interdisciplinary or untraditional approach to law ... .”).

  16. [16]. See infra Section II.C.

  17. [17]. See generally Richard H. Thaler, Anomalies: Saving, Fungibility, and Mental Accounts, 4 J. Econ. Persps. 193 (1990) (arguing that mental accounts prevent people from thinking about money as fungible across types of wealth); see infra Part III.

  18. [18]. See infra Section III.A.

  19. [19]. See infra notes 136–140 and accompanying text.

  20. [20]. See Kaplow & Shavell, supra note 8, at 33–36; N. Gregory Mankiw, Matthew Weinzierl & Danny Yagan, Optimal Taxation in Theory and Practice, 23 J. Econ. Persps. 147, 159–61 (2009).

  21. [21]. See Steven M. Sheffrin, Tax Fairness and Folk Justice 118–33 (2013); MatthewWeinzierl, Popular Acceptance of Inequality Due to Innate Brute Luck and Support for Classical Benefit-Based Taxation, 155 J. Pub. Econ. 54, 56–61 (2017); see infra Section III.B.

  22. [22]. To the extent that taxation based on desert in pretax income is a mistake, this is a “second-best” approach. See R.G. Lipsey & Kelvin Lancaster, The General Theory of Second Best,24 Rev. Econ. Stud.11, 12 (1956) (explaining that, if one part of the best policy outcome is not feasible, then the next-best solution may involve adopting policies that would otherwise not be optimal).

  23. [23]. This term was introduced by George H.W. Bush in his speech accepting the Republican presidential nomination, where he argued that common goals should be accomplished through the actions of America’s diverse array of community and civic organizations, which represent “a thousand points of light in a broad and peaceful sky.” George Bush, Address Accepting the Presidential Nomination at the Republican National Convention in New Orleans, Am. Presidency Project (Aug. 18, 1988), []. The Article means to use the term to convey the availability of many areas of the law, rather than an emphasis on voluntarism.

  24. [24]. See Jonathan Gruber,Public Finance and Public Policy 594–95 (4th ed. 2013) (explaining that the marginal cost of deviating from the optimal outcome increases in the size of the deviation from the optimal outcome).

  25. [25]. Coronavirus Aid, Relief, and Economic Security (“CARES”) Act, Pub. L. No. 116-136,
    § 2201, 134 Stat. 281, 335 (2020) (providing tax refunds of $1,200 to individuals with incomes up to $75,000 as a part of COVID-19 economic relief).

  26. [26]. See infra Section III.C.

  27. [27]. See, e.g., Ronald M. Dworkin, Is Wealth a Value?, 9 J. Legal Stud. 191, 217–226 (1980); C. Edwin Baker, The Ideology of the Economic Analysis of Law, 5 Phil. & Pub. Affs. 3, 16–19 (1975).

  28. [28]. See, e.g., Dworkin, supra note 27, at 202. Nor is the goal describing “how policies can shape political preferences about redistribution.” Gillian Lester, Can Joe the Plumber Support Redistribution? Law, Social Preferences, and Sustainable Policy Design, 64 Tax L. Rev. 313, 318 (2011).

  29. [29]. See infra Section V.A.

  30. [30]. But see Lee Anne Fennell & Richard H. McAdams, The Distributive Deficit in Law and Economics, 100 Minn. L. Rev. 1051, 1097 (2016) (briefly discussing fairness preferences).

  31. [31]. See Mankiw, supra note 11, at 145.

  32. [32]. See id.

  33. [33]. See Oren Cass, Economic Piety Is a Crisis for Workers, Atlantic (Nov. 27, 2018), https:// [] (describing the emphasis on economic growth in policymaking outside of taxes and transfers).

  34. [34]. Kaplow & Shavell, supra note 9, at 667. Even more stringently, the taxes need to be on labor income. See generally Alan J. Auerbach & James R. Hines Jr., Taxation and Economic Efficiency, in 3 Handbook of Public Economics 1347 (Alan J. Auerbach & Martin Feldstein eds., 2002) (describing the Atkinson-Stiglitz result that, under a certain set of assumptions, there should be no taxes on capital).

  35. [35]. Though this is the dominant view, there have been a variety of internal critiques to this view as well. See Zachary Liscow, Note, Reducing Inequality on the Cheap: When Legal Rule Design Should Incorporate Equity as Well as Efficiency, 123 Yale L.J. 2478, 2481 (2013) (describing how redistribution through legal rules can be more efficient); Chris William Sanchirico, Taxes Versus Legal Rules as Instruments for Equity: A More Equitable View, 29 J. Legal Stud. 797, 805–06 (2000) (similar). This Article differs from earlier work in that these works accepted the basic framework of finding the most efficient opportunities for redistribution among all possible choices. This Article, in contrast, rejects that framework as conflicting with social reality.

  36. [36]. See Lewis A. Kornhauser, Wealth Maximization, in 3 The New Palgrave Dictionary of Economics and the Law 679–84 (Peter Newman ed., 1998) (describing Kaldor-Hicks efficiency); Liscow, supra note 15, at 1652 (same).

  37. [37]. Dep’t of Transp. Memo, supra note 5, at 15.

  38. [38]. Kaplow & Shavell, supra note 9, at 668-69; Louis Kaplow & Steven Shavell, Any Non-Welfarist Method of Policy Assessment Violates the Pareto Principle, 109 J. Pol. Econ. 281, 284 (2001).

  39. [39]. See, e.g., Alan Greenspan et al., An Open Letter, Greg Mankiw’s Blog (Mar. 5, 2015), [] (writing an open letter from more than a dozen prominent economists to congressional leaders arguing that “[t]rade is beneficial for our society as a whole, but the benefits are unevenly distributed” yet “economy-wide benefits resulting from increased trade provide resources[,]” which can be used to “help[] those who are adversely affected”); Robert Whaples, The Policy Views of American Economic Association Members: The Results of a New Survey, 6 Econ J. Watch 337, 340 tbl.1 (2009) (finding support among economists for the position that the United States should continue to liberalize trade and increase support for affected workers).

  40. [40]. Gruber, supra note 24, at 328–34.

  41. [41]. Janet Currie & Firouz Gahvari, Transfers in Cash and In-Kind: Theory Meets the Data, 46 J. Econ. Literature 333, 341–42 (2008).

  42. [42]. Kaplow & Shavell, supra note 8, at 24–27.

  43. [43]. See Andreu Mas-Colell, Michael D. Whinston & Jerry R. Green, Microeconomic Theory 117–22, 817–50 (1995); Lewis Kornhauser, The Economic Analysis of Law, Stan. Encyclopedia Phil. (July 17, 2017), [] (“The evaluative tradition in economics is resolutely welfarist.”).

  44. [44]. See generally Sarah B. Lawsky, On the Edge: Declining Marginal Utility and Tax Policy, 95 Minn. L. Rev. 904 (2011) (discussing the declining marginal utility of income).

  45. [45]. See generally Lily L. Batchelder, Optimal Tax Theory as a Theory of Distributive Justice (Aug. 8, 2020) (unpublished manuscript),
    _id=3724691 [] (describing the similarity between optimal tax theory and other egalitarian theories of justice).

  46. [46]. Kaplow & Shavell, supra note 8, at 27, 30–31.

  47. [47]. Or like water from a leaky bucket. See Arthur M. Okun, Equality and Efficiency: The Big Tradeoff 89 (2015).

  48. [48]. The traditional view is to have low, possibly even zero, tax rates on capital income. See generally Mankiw et al., supra note 20 (discussing lessons from tax theory for tax policy). That view is in flux. See Peter Diamond & Emmanuel Saez, The Case for a Progressive Tax: From Basic Research to Policy Recommendations, 25 J. Econ. Persps. 165, 177–83 (2011) (discussing reasons to have at least a tax of greater than zero on capital income). This Article treats capital income tax as falling under the category of nontax legal rules, since it is not a tax on labor income.

  49. [49]. Liscow, supra note 15, at 1664.

  50. [50]. See generally J.A. Mirrlees, The Theory of Optimal Taxation, in 3 Handbook of Mathematical Economics 1197 (Kenneth J. Arrow & Michael D. Intriligator eds., 1986) (describing the classic treatment).

  51. [51]. See Richard H. Thaler, Behavioral Economics: Past, Present, and Future, 106 Am. Econ. Rev. 1577, 1578 (2016).

  52. [52]. Kaplow & Shavell, supra note 9, at 667.

  53. [53]. F. Patrick Hubbard, The Nature and Impact of the “Tort Reform” Movement, 35 Hofstra L. Rev. 437, 473–74 (2006).

  54. [54]. Id. at 473.

  55. [55]. Id. at 454.

  56. [56]. Id. at 475.

  57. [57]. Jean R. Sternlight, Panacea or Corporate Tool?: Debunking the Supreme Court’s Preference for Binding Arbitration, 74 Wash. U. L.Q. 637, 684 (1996) (listing ways in which poorer plaintiffs are procedurally and monetarily disadvantaged by arbitration over jury trial).

  58. [58]. See generally AT&T Mobility LLC v. Concepcion, 563 U.S. 333 (2011) (holding the Federal Arbitration Act pre-empts California law regarding collective-arbitration waivers).

  59. [59]. Erwin Chemerinsky, Closing the Courthouse Doors, 90 Denv. U. L. Rev. 317, 319 (2012); see also Judith Resnik, Diffusing Disputes: The Public in the Private of Arbitration, the Private in Courts, and the Erasure of Rights, 124 Yale L.J. 2804, 2891 (2015) (discussing the majority opinion in the AT&T case).

  60. [60]. Jedidiah Britton-Purdy, David Singh Grewal, Amy Kapczynski & K. Sabeel Rahman, Building a Law-and-Political-Economy Framework: Beyond the Twentieth-Century Synthesis, 129 Yale L.J. 1784, 1801–06 (2020) (describing trajectories toward efficiency and away from equity across many areas of the law).

  61. [61]. Exec. Order No. 12,291, 46 Fed. Reg. 13,193 (Feb. 17, 1981). Later executive orders do mention “equity” and “distributive impacts.” Richard L. Revesz, Regulation and Distribution,93 N.Y.U. L. Rev. 1489, 1541 (2018). However, a review of rulemakings by Ricky Revesz shows that these have not affected agency behavior. Id. at 1534­–43.

  62. [62]. See Henry Hansmann & Reinier Kraakman, The End of History for Corporate Law, 89 GEO. L.J. 439, 439 (2001) (“There is no longer any serious competitor to the view that corporate law should principally strive to increase long-term shareholder value.”).

  63. [63]. Robert H. Bork, Legislative Intent and the Policy of the Sherman Act, 9 J.L. & Econ.7, 7 (1966) (arguing that the sole goal of the Sherman Act is efficiency).

  64. [64]. Robert H. Lande, A Traditional and Textualist Analysis of the Goals of Antitrust: Efficiency, Preventing Theft from Consumers, and Consumer Choice, 81 Fordham L. Rev. 2349, 2354 (2013). Note that technically the focus was efficiency to consumers, but in any case, the focus was efficiency. Daniel A. Crane, The Tempting of Antitrust: Robert Bork and the Goals of Antitrust Policy, 79 Antitrust L.J. 835, 835 (2014).

  65. [65]. See infra Section V.B.

  66. [66]. See, e.g., Roberto Mangabeira Unger, What Should Legal Analysis Become? 9 (1996); Baker, supra note 27, at 16–19; Dworkin, supra note 27, at 194; Richard S. Markovits, Why Kaplow and Shavell’s “Double-Distortion Argument” Articles Are Wrong, 13 Geo. Mason L. Rev. 511, 529 (2005).

  67. [67]. Jacob S. Hacker, The Institutional Foundations of Middle-Class Democracy, 6 Pol’y Network 33, 35 (2011) (coining the term).

  68. [68]. See infra notes 136–40 and accompanying text.

  69. [69]. Thaler, supra note 17, at 194–95.

  70. [70]. Id. at 194.

  71. [71]. Id. at 195.

  72. [72]. See supra Section II.A.

  73. [73]. G.A. Res. 217 (III) A, Universal Declaration of Human Rights, at 25 (Dec. 10, 1948).

  74. [74]. See Larry Bye & Alyssa Ghirardelli, NORC at U. Chi., Topline Report: American Health Values Survey Data Tables 5 (2016),
    reports/2016/rwjf437263 [] (finding that an overwhelming majority of Americans believe that “[e]nsuring that low-income Americans have the same chance to get good quality health care as those who are better off financially” should be a “top” or “high” priority).

  75. [75]. See Currie & Gahvari, supra note 41, at 335–37 (describing the in-kind programs across countries).

  76. [76]. See Gruber, supra note 24, at 294-98.

  77. [77]. Id. at 366.

  78. [78]. Consider someone who earns $20,000 per year. For Medicaid beneficiaries ages 19-64, average annual spending per enrollee was $9,079 in 2010. David Lassman, Micah Hartman, Benjamin Washington, Kimberly Andrews & Aaron Catlin, US Health Spending Trends by Age and Gender: Selected Years 2002–10, 33 Health Affs. 815, 820 exhibit 4 (2014). It seems likely that someone so poor would prefer the cash to the full $9,079 in healthcare.

  79. [79]. See generally Douglas A. Kysar, Preferences for Processes: The Process/Product Distinction and the Regulation of Consumer Choice, 118 Harv. L. Rev.526 (2004) (arguing for the centrality of non-standard preferences for process, not just products, in the context of consumer choice).

  80. [80]. See Richard L. Revesz & Michael A. Livermore, Retaking Rationality: How Cost-Benefit Analysis Can Better Protect the Environment and Our Health 83–84 (2008) (arguing that “a cost-benefit analysis using an average value of a statistical life is unmoored from its economic justification”).

  81. [81]. See James A. Henderson, Jr., Richard N. Pearson, Douglas A. Kysar & John A. Siliciano, The Torts Process 35–39 (7th ed. 2007) (describing how major theories of tort law, including instrumental, corrective justice, and civil recourse, do not focus on redistribution from the rich to the poor for its own sake).

  82. [82]. See generally Conor Clarke & Edward Fox, Perceptions of Taxing and Spending: A Survey Experiment, 124 Yale L.J. 1252 (2015) (describing the apparent ease of providing tax breaks instead of economically equivalent direct transfers).

  83. [83]. Id. at 1277 tbl.4.

  84. [84]. See generally Zachary Liscow & Abigail Pershing, Why Is So Much Redistribution In-Kind and Not in Cash? Evidence from a Survey Experiment (Yale L. & Econ., Working Paper, 2020), https:// [] (finding in an online survey experiment greater support for cash than for in kind redistribution of the same value).

  85. [85]. Id. at 6–8.

  86. [86]. Id.

  87. [87]. Id.

  88. [88]. Id. at 14–15.

  89. [89]. See generally Martin Gilens, Affluence and Influence: Economic Inequality and Political Power in America (2012) (discussing the link between wealth and political influence).

  90. [90]. Edward J. McCaffery & Jonathan Baron, The Humpty Dumpty Blues: Disaggregation Bias in the Evaluation of Tax Systems, 91 Organizational Behav. & Hum. Decision Processes 230, 230 (2003).

  91. [91]. Id. at 240–41.

  92. [92]. This could be akin to the different “spheres of justice” that Michael Walzer describes, each with its own notion of the proper role of the state in adjudicating distributive issues. Michael Walzer, Spheres of Justice: A Defense of Pluralism and Equality 6–10 (1983).

  93. [93]. Liscow & Pershing, supra note 84, at 16–19.

  94. [94]. This psychology may be reinforced by the “endowment effect,” the finding that people are less willing to part with something they own than to acquire the same thing that they do not own. Daniel Kahneman, Jack L. Knetsch & Richard H. Thaler, Experimental Tests of the Endowment Effect and the Coase Theorem,98 J. Pol. Econ. 1325, 1326, 1328 (1990).

  95. [95]. See generally Liam Murphy & Thomas Nagel, The Myth of Ownership: Taxes and Justice (2002) (discussing the ethical sides of tax public policy).

  96. [96]. Id. at 35–36. Murphy and Nagel use somewhat different terminology than this Article uses. They distinguish between the normative concept of “pre-institutional desert” and the sociological or empirical concept of “entitlement,” also known as “institutional desert,” or how our institutions in fact generate deservingness. See Fred Feldman & Brad Skow, Desert, Stan. Encyclopedia Phil. (Sept. 22, 2020), [
    H8KE-QYB4]. Since this Article is primarily interested in describing people’s views on tax policy, rather than investigating where they come from, it combines the two ideas under the heading of “desert.”

  97. [97]. Murphy & Nagel, supra note 95, at 36.

  98. [98]. Id.

  99. [99]. Id.

  100. [100]. Id. at 8–9.

  101. [101]. Id. at 34.

  102. [102]. Id. at 175.

  103. [103]. Id. at 176.

  104. [104]. Id.

  105. [105]. Id. at 175.

  106. [106]. Mankiw et al., supra note 20, at 147 (summarizing the literature); Diamond & Saez, supra note 48, at 165 (disagreeing in some respects, but establishing many points of consensus).

  107. [107]. See generally Joseph Bankman & Thomas Griffith, Social Welfare and the Rate Structure: A New Look at Progressive Taxation, 75 Calif. L. Rev. 1905 (1987) (an example of tax scholars discussing optimal taxation).

  108. [108]. See, for example, the debate between Mankiw et al. on one side and Diamond and Saez on the other. Mankiw et al., supra note 20, at 155–59; Diamond & Saez, supra note 48, at 175–77.

  109. [109]. Mankiw et al., supra note 20, at 155–59; Diamond & Saez, supra note 48, at 175–77.

  110. [110]. Emmanuel Saez, Optimal Income Transfer Programs: Intensive Versus Extensive Labor Supply Responses, 117 Q.J. Econ. 1039, 1059–66 (2002).

  111. [111]. Milton Friedman, Capitalism and Freedom 190–94 (1962); Daniel P. Moynihan, The Politics of a Guaranteed Income: The Nixon Administration and the Family Assistance Plan 50, 131 (1973); Robert Moffitt, Welfare Work Requirements with Paternalistic Government Preferences, 116 Econ. J. F441, F441 (2006).

  112. [112]. Saez, supra note 110, at 1061 tbl.2 panel B (middle row) (showing a guaranteed income of $7,300 in 1996 dollars). If anything, this demogrant estimate is small, given recent estimates. See Raj Chetty, Bounds on Elasticities with Optimization Frictions: A Synthesis of Micro and Macro Evidence on Labor Supply, 80 Econometrica 969, 1008 tbl.2 (2012) (showing that the average estimate of the extensive margin elasticity is 0.25, which would imply an even larger demogrant than the one described in the text above); see also Mankiw et al., supra note 20, at 158 (providing calibrations showing an even larger demogrant “equal to just over 60 percent of average income per worker in the economy”). Importantly, this analysis does not include “tagging,” allowing larger transfers to those, like the disabled, who have lower earnings abilities; incorporating tagging would lower the size of the demogrant. Nor does it include the possibility of provision of services like healthcare, which would also presumably lower the size of the optimal demogrant. See also Batchelder, supra note 45, at 20 (describing how similar estimates have been around since the beginning of the Mirrlees model).

  113. [113]. Probably the closest in the United States is the Alaska Permanent Fund, which used revenue from oil drilling to fund a “dividend” that averaged $1,300 per resident per year between 2009 and 2018. See Summary of Dividend Applications & Payments, Alaska Dep’t of Revenue, []. Whether something similar would be feasible in the rest of the country without significant oil wealth is an open question. The U.S. Internal Revenue Code does not provide transfers at all to those who do not work, even those with children. For those who earn no money, none of the (per child) $2,000 Child Tax Credit is available. I.R.C. § 24(d)(1)(B) (2018). Nor is the Earned Income Tax Credit available, as that also requires earned income. I.R.C. § 32 (2018). Temporary Assistance for Needy Families benefits have work requirements and typically have time limits.

  114. [114]. See supra note 113 and accompanying text.

  115. [115]. An interesting question is what other countries do, as that may suggest what is possible in the United States. Some Gulf States have an “implicit government job guarantee” for nationals, but that is different from a demogrant because it involves work. Steffen Hertog, The GCC’s National Employment Challenge, Wash. Post (July 31, 2014),
    news/monkey-cage/wp/2014/07/31/the-gccs-national-employment-challenge [
    FWB-4M8L]. Scandinavian countries have generous unemployment programs as well. For example, unemployment insurance in Denmark allows beneficiaries to receive up to 90 percent of their previous salary. However, receipt of unemployment benefits requires previous work experience (with exceptions for those who have just finished education or training) and active job searching, and a person is only entitled to benefits for 2 years within a 3-year period. Denmark – Unemployment Benefit, Eur. Comm’n,
    en&intPageId=4496 []. Again, this policy is not a demogrant because it is tied to work. Finally, the United Kingdom passed a small Child Trust Fund, in which children were given £500 from the government and the opportunity for parents and grandparents to save additional money tax-free, but these are far smaller than the sum of annual demogrants and in any case only go to children. See Child Trust Funds Act, (2004) § 3, 6 Current Law (U.K.); Zoe Williams, Why We Cannot Afford to Raid the Child Trust Fund Piggy Bank, Guardian (May 2, 2010, 1:21 PM), []; see also HM Revenue and Customs, Child Trust Fund Statistics: Detailed Distributional Analysis tbl.2 (Feb. 2013), https://assets. [] (showing contributions to Child Trust Funds in 2011 and 2012).

  116. [116]. Saez, supra note 110, at 1061 tbl.2 panel B (middle row).

  117. [117]. Id. In technical terms, the best way to raise money from middle and high-income earners is to have high tax rates on their “inframarginal” earnings, which are the dollars they earn that are far from their decision-making margin. For example, if a worker makes $40,000 per year and the government places high taxes on any earnings below $25,000, the worker is unlikely to cut back their hours in order to get under the $25,000 threshold.

  118. [118]. Id.

  119. [119]. The idea originally comes from George A. Akerlof, The Economics of “Tagging” as Applied to the Optimal Income Tax, Welfare Programs, and Manpower Planning, 68 Am. Econ. Rev. 8, 8 (1978).

  120. [120]. Mankiw et al., supra note 20, at 161–64; see Diamond & Saez, supra note 48, at 166.

  121. [121]. N. Gregory Mankiw & Matthew Weinzierl, The Optimal Taxation of Height: A Case Study of Utilitarian Income Redistribution, 2 Am. Econ. J.: Econ. Pol’y 155, 156 (2010).

  122. [122]. With advancements in genetics, an arguably dystopian future could be prescribed in which taxes are based not principally on earnings but instead on one’s genetic endowment.

  123. [123]. But see Batchelder, supra note 45, at 30–38 (arguing that such consequences need not be severe).

  124. [124]. Mankiw et al., supra note 20, at 163–64.

  125. [125]. Emmanuel Saez & Stefanie Stantcheva, Generalized Social Marginal Welfare Weights for Optimal Tax Theory, 106 Am. Econ. Rev. 24, 33 (2016) (presenting survey evidence suggesting that virtually no one supports tagging in their context).

  126. [126]. Mankiw et al., supra note 20, at 159–61; Diamond & Saez, supra note 48, app. at 189.

  127. [127]. Sheffrin, supra note 21, at 130–31.

  128. [128]. See generally Thomas Piketty & Emmanuel Saez, Income Inequality in the United States, 1913–1998, 118 Q.J. Econ. 1 (2003) (presenting the top shares of income and wages from 1913 to 1998). The extent of the rise is disputed. E.g., Richard V. Burkhauser, Shuaizhang Feng, Stephen P. Jenkins & Jeff Larrimore, Recent Trends in Top Income Shares in the United States: Reconciling Estimates from March CPS and IRS Tax Return Data, 94 Rev. Econ. & Stat. 371, 372 (2012). See generally Alan J. Auerbach, Laurence J. Kotlikoff & Darryl R. Koehler, U.S. Inequality and Fiscal Progressivity: An Intragenerational Accounting (Nat’l Bureau of Econ. Rsch., Working Paper No. 22032, 2016) (measuring income information within cohorts).

  129. [129]. See, e.g., Benjamin B. Lockwood & Matthew Weinzierl, Positive and Normative Judgments Implicit in U.S. Tax Policy, and the Costs of Unequal Growth and Recessions, 77 J. Monetary Econ. 30, 46 (2016) (showing that taxation has redistributed considerably less to the poor over time, even considering changing economic circumstances).

  130. [130]. Am. Enter. inst. for Pub. Pol’y Rsch., Public Opinion on Taxes: 1937 to Today 25 (Apr. 2012); Sheffrin, supra note 21, at 119–20.

  131. [131]. Rebbecca Reed-Arthurs & Steven M. Sheffrin, Windows into Public Attitudes Towards Redistribution, 102 Proc. of the Nat’l Tax Ass’n 165, 168 (2009).

  132. [132]. See generally Daniel Markovits, The Meritocracy Trap (2019) (arguing that an ethos of “meritocracy” pervades our age, including the idea that, because success is meritocratic, the earnings from career success are thus deserved).

  133. [133]. Martin Feldstein, The Mirrlees Review, 50 J. Econ. Lit. 781, 783 (2012).

  134. [134]. Id.

  135. [135]. See generally Kenneth Scheve & David Stasavage, Taxing the Rich: A History of Fiscal Fairness in the United States and Europe (2016) (making this argument).

  136. [136]. The method first considers the incentive effects of taxation: if people’s incomes are taxed, they will tend to work less. And then it asks how much people implicitly weight, or value, a dollar in the hands of a rich versus poor person, given those incentive effects and the taxes that we see. So, the more redistribution there is to the poor, despite the efficiency harms of doing so, the larger the implicit weight on the poor relative to the rich. See Nathaniel Hendren, Efficient Welfare Weights 30–31 (Aug. 2019) (unpublished manuscript),
    files/hendren/files/eww.pdf []; see also Lockwood & Weinzierl, supra note 129, at 46.

  137. [137]. Hendren, supra note 136, at 24 fig.6.

  138. [138]. These results are conservatively, and roughly, approximated by the “logarithm” function.

  139. [139]. For incomes, see PK, Average, Median, Top 1%, and Income Percentiles in 2018 in the United States, DQYDJ (Nov. 2, 2020), [].

  140. [140]. See, e.g., Raj Chetty, A New Method of Estimating Risk Aversion, 96Am. Econ. Rev. 1821, 1830 (2006) (finding a “coefficient of relative risk aversion” of approximately one, which has this implication). The “logarithmic” utility function implied by this analysis is also suggested by cross-country studies measuring stated happiness. See generally Néstor Gandelman & Rubén Hernández-Murillo, Risk Aversion at the Country Level, 97 Fed. Rsrv. Bank St. Louis Rev. 53 (2015) (finding that this utility function is implied by measures of happiness). Note that many other estimates imply even higher risk aversion, which suggest an even larger divergence in distributional weights between the rich and the poor. Chetty, supra, at 1821.

  141. [141]. For example, issues of interpersonal comparability arise in going from within-person to between-person comparisons.

  142. [142]. See, e.g., Raymond Geuss, Philosophy and Real Politics 36, 49–50 (2008) (describing how conceptual instruments used to analyze social reality are often inextricable from social reality itself).

  143. [143]. See, e.g., Philip Bump, Who Actually Wants Congressional Republicans to Pass Trump’s Tax Bill?, Wash. Post (Nov. 2, 2017),
    11/02/who-actually-wants-congressional-republicans-to-pass-trumps-tax-bill [https://perma.
    ]; Marisa Schultz, GOP’s Big Donors Threaten to Close Wallets if Tax Reform Isn’t Passed, N.Y. Post (Nov. 6, 2017, 4:35 PM), []; Dylan Scott, House Republican: My Donors Told Me to Pass the Tax Bill “or Don’t Ever Call Me Again”, Vox (Nov. 7, 2017, 11:30 AM), [].

  144. [144]. See, e.g., Alberto Alesina, Stefanie Stantcheva & Edoardo Teso, Intergenerational Mobility and Preferences for Redistribution,108 Am. Econ. Rev. 521, 521 (2018); Jimmy Charité, Raymond Fisman, & Ilyana Kuziemko, Reference Points and Redistributive Preferences: Experiment Evidence 1 (Nat’l Bureau of Econ. Rsch., Working Paper No. 21009, 2016); Raymond Fisman, Keith Gladstone, Ilyana Kuziemko & Suresh Naidu, Do Americans Want to Tax Wealth? Evidence from Online Surveys, 188J. Pub. Econ.1, 1–2 (2020); Matthew Weinzierl, A Welfarist Role for Nonwelfarist Rules: An Example with Envy 1–2 (Nat’l Bureau of Econ. Rsch., Working Paper No. 23587, 2017); Stefanie Stantcheva, Understanding Tax Policy: How Do People Reason?, 136 Q.J. Econ. 2309 (2021).

  145. [145]. Christopher Freiman & Shaun Nichols, Is Desert in the Details?,82 Phil. & Phenomenological Rsch.121, 127 (2011).

  146. [146]. N. Gregory Mankiw, Defending the One Percent, 27J. Econ. Persps. 21, 29 (2013).

  147. [147]. Weinzierl, supra note 21, at 54.

  148. [148]. Id. at 56.

  149. [149]. The interface clearly showed how much money each party ended up with given a proposed allocation, making outcomes salient. Id. at 58.

  150. [150]. Id. at 56.

  151. [151]. Sheffrin, supra note 21, at 3 (emphasis omitted).

  152. [152]. Zachary Liscow & Edward Fox, The Psychology of Taxing Capital Income: Evidence from a Survey Experiment on the Realization Rule 87 (Aug. 2021) (unpublished manuscript), [].

  153. [153]. President George W. Bush, President’s Radio Address, (Apr. 15, 2006), https:// [

  154. [154]. Latest Poll, FiveThirtyEight (Aug. 20, 2021, 5:29 PM),
    polls/president-primary-d []; 2020 Democratic Popular Vote, Real Clear Politics,
    _count.html [] (showing primary vote totals through May 2020).

  155. [155]. President Barack Obama said, “Understand we’ve never begrudged success in America. We aspire to it. We admire folks who start new businesses, create jobs, and invent the products that enrich our lives. And we expect them to be rewarded handsomely for it.” President Obama on Inequality (Transcript), Politico (Dec. 4, 2013, 1:54 PM),
    obama-income-inequality-100662 [].See Markovits, supra note 132, at 79.

  156. [156]. Meet the Candidates: 18. Does Anyone Deserve to Have a Billion Dollars?, N.Y. Times (2019), [].

  157. [157]. See Mankiw, supra note 146, at 32 (arguing for a “just deserts” goal on taxation, in which “people should receive compensation congruent with their contributions”); contra Murphy & Nagel, supra note 95, at 32–37 (arguing the opposite); see also Sheffrin, supra note 21, at 8–10 (describing “resonance” with existing policies).

  158. [158]. As Section V.C discusses, changing views should be considered, including by changing policies if those views do change.

  159. [159]. See Sheffrin, supra note 21, at 125–28 (finding those with incomes greater than $200,000 generally oppose using the tax system to redistribute, while those making less $25,000 strongly support doing so, and assembling other evidence); Alberto F. Alesina & Paola Giuliano, Preferences for Redistribution 27–35 (Nat’l Bureau of Econ. Rsch., Working Paper No. 14825, 2009), [] (finding income is negatively correlated with preferences for redistribution). Of course, the relationship between voters’ policy views and actual policymaking is unclear. For example, some argue that the policy preferences of none but the richest drive policy. See generally Gilens, supra note 89 (showing that, controlling for the views of the well-off, the views of the poor show no relationship with actual policy). Nevertheless, it seems reasonable to argue that widespread voters’ views—especially among the well-off, who are likely most inclined to value desert—have an important impact on those that they elect. Indeed, in Liscow and Pershing’s study, for example, the rich are especially skeptical of redistribution through cash. See Liscow & Pershing, supra note 84, at 32.

  160. [160]. See Frank I. Michelman, Property, Utility, and Fairness: Comments on the Ethical Foundations of “Just Compensation” Law,80 Harv. L. Rev.1165, 1172 (1967).

  161. [161]. Recall that efficient policies are based on willingness to pay. We can often infer willingness to pay by observing what people actually choose—and are thus willing—to pay for a given good or service. See Section II.A.

  162. [162]. See Liscow, supra note 15, at 1655 (discussing efficiency and neutrality).

  163. [163]. Britton-Purdy et al., supra note 60, at 1832.

  164. [164]. Benjamin I. Page & Robert Y. Shapiro, Effects of Public Opinion on Policy, 77 Am. Pol. Sci. Rev. 175, 188–89 (1983) (finding, in a classic study, that Congress is responsive, though imperfectly so, to public attitudes); see Gilens, supra note 89, at 234 (finding Congress responsive at least to the attitudes of the wealthy).

  165. [165]. Liscow, supra note 15, at 1680 (discussing when efficiency is and is not neutral and describing how it tends to provide larger legal entitlements to the rich). See generally Dworkin, supra note 27 (providing a prominent early example of non-neutrality).

  166. [166]. Liscow, supra note 15, at 1680–81.

  167. [167]. Pew Rsch. Ctr., supra note 2, at 18; See Rampell, supra note 3.

  168. [168]. For example, though some would disagree with the prescribed degree of redistribution, the Rawlsian goal of achieving the maximum benefit for the least well-off is broadly consistent with the normative approach in this paper, as it is still concerned with helping the less well-off under resource constraints. John Rawls, A Theory of Justice 258–65 (1971). Rawls also prescribes equal basic liberties, which is consistent with the approach: the parsimonious goal of addressing tradeoffs among limited resources is just a partial one and does not defeat the many other important background principles. Id. at 195–201.

  169. [169]. These are the values implicit, for example, in typical optimal taxation papers. See, e.g., Saez, supra note 110, at 1044–56.

  170. [170]. Alexander Hertel-Fernandez, Roosevelt Inst., How Policymakers Can Craft Measures That Endure and Build Political Power, 4 (2020),
    wp-content/uploads/2020/06/RI_How-Policymakers-Can-Craft-Measures-that-Endure-and-Build-Political-Power-Working-Paper-2020.pdf [].

  171. [171]. See supra Section III.B.

  172. [172]. Exec. Order No. 13,985, 86 Fed. Reg. 7,009 (Jan. 20, 2021) (discussing advancing equity); Modernizing Regulatory Review, 86 Fed. Reg. 7223 (Jan. 20, 2021) (requiring “procedures that take into account the distributional consequences of regulations”).

  173. [173]. Hendren, supra note 136, at 2.

  174. [174]. Bouchard, supra note 7.

  175. [175]. Dep’t of Transp. Memo, supra note 5.

  176. [176]. See supra Section III.B.

  177. [177]. See, e.g., Rawls, supra note 168, at 359–60.

  178. [178]. Gruber, supra note 24, at 594–95. Those familiar with economics may know this as the “envelope theorem.”

  179. [179]. Id. at 595–96.

  180. [180]. See David Gamage, How Should Governments Promote Distributive Justice?: A Framework for Analyzing the Optimal Choice of Tax Instruments, 68 Tax L. Rev. 1, 11 (2014) (making a related point about how imperfections in taxation mean that a variety of tax instruments should be used).

  181. [181]. Framework statutes moderate separation of powers and delegations of power, often between the legislature and executive, though the term can also be read broadly to encompass procedural meta-rules that govern the production of rules. Ernest A. Young, Toward a Framework Statute for Supranational Adjudication, 57 Emory L.J. 93, 98 (2007). For example, they can be used for budget processes. Kenneth W. Dam, The American Fiscal Constitution, 44 U. Chi. L. Rev. 271, 279–80 (1977); Elizabeth Garrett, The Purposes of Framework Legislation, 14 J. Contemp. Legal Issues 717, 724 (2005). Other examples are agency organic statutes. Young, supra, at 98. Ackerman has proposed a framework statute to govern the president’s emergency powers. See generally Bruce Ackerman, Before The Next Attack: Preserving Civil Liberties In An Age of Terrorism (2006) (discussing the idea of an “emergency constitution” that temporarily takes effect during times of crisis).

  182. [182]. Indeed, the United Kingdom provides such weights in its “Green Book,” though without accounting for the redistribution that already takes place. HM Treasury, The Green Book: Central Government Guidance on Appraisal and Evaluation 9 (2020), https://assets. [].

  183. [183]. See Revesz, supra note 61, at 1534–43 (describing how current regulations have failed to attend to distributional concerns).

  184. [184]. The Social Cost of Carbon: Estimating the Benefits of Reducing Greenhouse Gas Emissions, EPA, [

  185. [185]. U.S. Dep’t of Lab., Conformity Requirements for State UI Laws: Unemployment Insurance Extended Benefits 1–2, [].

  186. [186]. See infra Section V.C for more on this issue.

  187. [187]. Another way this could arise is from the education that happens through policymaking. For example, congestion pricing schemes have sometimes faced opposition that then subsides after they are implemented. See, e.g., Björn Hårsman & John M. Quigley, Political and Public Acceptability of Congestion Pricing: Ideology and Self-Interest, 29J. Pol’y Analysis & Mgmt.854, 855 (2010). So, the less of one type of redistribution we have, the less people learn about it, further reducing its use. Indeed, the political process is often where policy views are adjudicated—and, indeed, often where they are formed in the first place. See, e.g., Reva B. Siegel, Constitutional Culture, Social Movement Conflict and Constitutional Change: The Case of the De Facto ERA, 94 Calif. L. Rev. 1323, 1341–42 (2006).

  188. [188]. Zachary Liscow, Are Court Orders Sticky? Evidence on Distributional Impacts from School Finance Litigation,15 J. Empirical Legal Stud.4, 18–26 (2018). But see Richard T. Boylan & Naci Mocan, Intended and Unintended Consequences of Prison Reform, 30 J.L. Econ. & Org. 558, 561 (2014) (showing evidence from a much smaller change—mandating more spending on prisoners—suggesting that social welfare spending was cut to pay for it).

  189. [189]. See, e.g., Robert Cooter, Prices and Sanctions, 84 Colum. L. Rev. 1523, 1532 (1984).

  190. [190]. For example, Finland has income-based fines. Joe Pinsker,Finland, Home of the $103,000 Speeding Ticket, Atlantic (Mar. 12, 2015),

    03/finland-home-of-the-103000-speeding-ticket/387484 [].
  191. [191]. Amy Finkelstein, Nathaniel Hendren & Erzo F.P. Luttmer, The Value of Medicaid: Interpreting Results from the Oregon Health Insurance Experiment, 127 J. Pol. Econ. 2836, 2839–40 (2019) (providing some estimates that Medicaid recipients are willing to pay considerably less than the cost of provision).

  192. [192]. See Gary T. Schwartz, Mixed Theories of Tort Law: Affirming Both Deterrence and Corrective Justice, 75 Tex. L. Rev. 1801, 1807–08 (1997); Judith Jarvis Thomson, The Trolley Problem, 94 Yale L.J. 1395, 1409 (1985) (describing conflicting moral intuitions about sacrificing one person to save five, depending upon how the one is sacrificed).

  193. [193]. Lipsey & Lancaster, supra note 22, at 31–32.

  194. [194]. See supra Part II.

  195. [195]. See supra Part III.

  196. [196]. See supra Sections II.A–II.B.

  197. [197]. See, e.g., David A. Weisbach, Distributionally Weighted Cost-Benefit Analysis: Welfare Economics Meets Organizational Design,7 J. Legal Analysis151, 177–78 (2015) (pointing to “a version of a political Coase theorem” in which these tax changes occur). But see Matthew D. Adler & Eric A. Posner,New Foundations of Cost-Benefit Analysis 130–31, 142–46, 152 (2006) (discussing adjustments to willingness to pay to compensate for parties’ different incomes, but without rooting it in a particular political economy); Revesz, supra note 61, at 1578 (arguing for the establishment of a body to consider the distributional consequences of regulations on a case-by-case basis, but explicitly not considering general distributional consequences).

  198. [198]. See Adam M. Samaha, Death and Paperwork Reduction, 65 Duke L.J. 279, 323–26 (2015).

  199. [199]. See, e.g., Kate Andrias, The New Labor Law, 126 Yale L.J. 2, 70 (2016).

  200. [200]. See generally Lina Khan & Sandeep Vaheesan, Market Power and Inequality: The Antitrust Counterrevolution and Its Discontents, 11 Harv. L. & Pol’y Rev. 235 (2017) (arguing that stronger antitrust enforcement would reduce inequality).

  201. [201]. See generally David H. Autor, Alan Manning & Christopher L. Smith, The Contribution of the Minimum Wage to US Wage Inequality over Three Decades: A Reassessment, 8 Am. Econ. J.: Applied Econ. 58 (2016) (finding that minimum wage has reduced income inequality).

  202. [202]. See generally Am. Med. Ass’n, Caps on Damages (2017), [
    HX2D-YEBS] (discussing caps on non-economic damages); see also Jonathan D. Glater, To the Trenches: The Tort War is Raging On, N.Y. Times (June 22, 2008),
    22/business/22tort.html [] (discussing changes in tort law).

  203. [203]. See generally Lucinda M. Finley, The Hidden Victims of Tort Reform: Women, Children, and the Elderly, 53 Emory L.J. 1263 (2004) (arguing that damages caps had a disproportionate impact on women, children, the elderly, and the disabled, since juries had been implicitly using categories like pain and suffering to make up for the fact that actual damages awards for these categories—being based on lost income—often were disproportionately low).

  204. [204]. See, e.g., Martha Chamallas, Civil Rights in Ordinary Tort Cases: Race, Gender, and the Calculation of Economic Loss, 38 Loy. L.A. L. Rev. 1435, 1450–51 (2005).

  205. [205]. Victim Compensation Fund Frequently Asked Questions, U.S. Dep’t Just. (Apr. 13, 2004), https:// [].

  206. [206]. Derek Thompson,Why Child Care Is So Ridiculously Expensive,Atlantic (Nov. 26, 2019), [] (“In California, the cost of a typical day-care center is now equal toalmost halfof the median income of a single mother.”).

  207. [207]. See Jacob Goldstein, How the Poor, The Middle Class and the Rich Spend Their Money, NPR (Aug. 1, 2012), [] (showing that the poor overwhelmingly spend their money on such “necessities”).

  208. [208]. Liscow & Pershing, supra note 84, at 5.

  209. [209]. Currie & Gahvari, supra note 41, at 333.

  210. [210]. See 2019 Public Opinion Polling on Housing Affordability, Opportunity Starts at Home (Mar. 28, 2019), [
    3QZD-MRWZ] (finding that 85 percent of respondents believe that “ensuring everyone has a safe, decent, affordable place to live should be a ‘top national priority’” and that 78 percent of respondents “say [that] government has an important role to play in making sure there are enough affordable places [for everyone] to live”).

  211. [211]. See generally Vanessa S. Williamson, Read My Lips: Why Americans Are Proud to Pay Taxes (2017) (detailing interviews about how individuals feel about paying taxes).

  212. [212]. I.R.C. § 3111(a)–(b) (2018) (imposing taxes on employers based on employee wages for hospital and old age and disability insurance).

  213. [213]. See I.R.C. §§ 3301–11 (2018) (imposing unemployment taxes on employers and setting out the approval process for state unemployment systems).

  214. [214]. This is because the payroll tax rate paid by employers has increased from 1 percent to 7.65 percent from the 1940s to today. Social Security & Medicare Tax Rates, Soc. Sec. Admin., []. And estimates suggest that such tax changes are largely reflected in reduced wages. See, e.g., Jonathan Gruber, The Incidence of Payroll Taxation: Evidence from Chile, 15 J. Lab. Econ. S72, S99–100 (1997) (finding such wage impacts).

  215. [215]. At the same time, means-testing the payments for social insurance programs could cause its own issues, possibly undercutting support for such programs.

  216. [216]. For tax policy implications when pretax income has moral weight, see Steven M. Sheffrin, What Role Can Desert Play in Designing Tax Policies?, 15 Pitt. Tax Rev. 137, 152–60 (2017). This Article is agnostic on that question. See also Ilan Benshalom, Recalibrating Moral Feasibility Boundaries of Taxation, 74 Tax L. Rev. (forthcoming 2021) (generally discussing tax policy in light of public attitudes).

  217. [217]. Gruber, supra note 24, at 559–61.

  218. [218]. See I.R.C. § 162(m) (2018) (limiting the deductibility of certain employees’ salaries at publicly held corporations).

  219. [219]. McCaffery & Baron, supra note 90, at 235–36 (showing support for taxing businesses).

  220. [220]. Mankiw et al., supra note 20, at 167–68 (discussing why taxes on capital should be zero); Diamond & Saez, supra note 48, at 180 (discussing why capital should be taxed, albeit noting “efficiency cost[s]”).

  221. [221]. I.R.C. § 1(h) (establishing preferential tax rates on capital gains and dividends).

  222. [222]. Id. § 1(a)–(e) (providing income tax rates).

  223. [223]. See, e.g., David Kamin, Wash. Ctr. for Equitable Growth, Taxing Capital: Paths to a Fairer and Broader U.S. Tax System 23 (2016),
    uploads/2016/08/081016-kamin-taxing-capital.pdf [].

  224. [224]. McCaffery & Baron, supra note 90, at 231–32; Steven Sheffrin & Rujun Zhao, Public Perceptions of the Tax Avoidance of Corporations and the Wealthy, 61 Empirical Econ. 259, 268 (2021); Liscow & Fox, supra note 152, at 38–40.

  225. [225]. CBS News & N.Y. Times, Americans’ Views on Taxes: January 20-23, 2012, at 1 (Jan. 24, 2012, 12:00 PM), [

  226. [226]. Ilyana Kuziemko, Michael I. Norton, Emmanuel Saez & Stefanie Stantcheva,How Elastic Are Preferences for Redistribution? Evidence from Randomized Survey Experiments, 105 Am. Econ. Rev. 1478, 1478–83 (2015) (trying to measure these preferences).

  227. [227]. See Quick Charts: Value-Added Tax (VAT) Rates, PWC,
    ID/Value-added-tax-(VAT)-rates [] (showing taxes across countries).

  228. [228]. Gilbert E. Metcalf, Value-Added Taxation: A Tax Whose Time Has Come?, 9 J. Econ. Persps. 121, 133–34 (1995) (describing the regressivity of the tax, but also noting some debate on the topic).

  229. [229]. Kyle Pomerleau, How Scandinavian Countries Pay for Their Government Spending, Tax Found. (June 10, 2015), [] (describing the Nordic model); Income Inequality, OECD, [
    Z3TV-7PR3] (similar).

  230. [230]. See, e.g., Clayton P. Gillette,Local Redistribution, Living Wage Ordinances, and Judicial Intervention, 101 Nw. U. L. Rev. 1057, 1057–67 (2007)(summarizing the traditional criticism).

  231. [231]. See generally Gilens, supra note 89 (discussing the link between income and political influence); Nancy MacLean, Democracy in Chains: The Deep History of the Radical Right’s Stealth Plan for America (2017) (discussing wealthy individuals’ influence on the political system).

  232. [232]. Scheve & Stasavage, supra note 135, at 42–46.

  233. [233]. For this paragraph, see supra notes 77–78 and accompanying text.

  234. [234]. Robert Pear&Thomas Kaplan, G.O.P. Plans to Replace Health Care Law With ‘Universal Access’, N.Y. Times (Dec. 15, 2016), [] (describing House Republicans’ plan of achieving “universal access” to quality, affordable healthcare).

  235. [235]. See, e.g., Pew Rsch. Ctr., supra note 2, at 18.

  236. [236]. See, e.g., Rampell, supra note 3.

  237. [237]. See Pear & Kaplan, supra note 234; Bye & Ghirardelli, supra note 74, at 11.

  238. [238]. Frank Newport, Americans Continue to Say U.S. Wealth Distribution is Unfair, Gallup (May 4, 2015), [].

  239. [239]. See, e.g., Alex Raskolnikov, Distributional Arguments, in Reverse, 105 Minn. L. Rev. 1583, 1624–31 (2021) (describing compensation failures); Aanund Hylland & Richard Zeckhauser, Distributional Objectives Should Affect Taxes but Not Program Choice or Design, 81 Scandinavian J. Econ. 264, 281 (1979) (using the tax system to compensate for distributional impacts); Weisbach, supra note 197, at 161 (“[W]e can replace the redistributive regulations with efficient ones and make a corresponding adjustment to the income tax. We keep the redistributive effects but eliminate the inefficiency.”); Liscow, supra note 35, at 2507 n.60 (describing a lack of compensation in federal cost-benefit analysis).

  240. [240]. Pew Rsch. Ctr., supra note 2, at 18.

  241. [241]. David Epstein & Sharyn O’Halloran, Delegating Powers: A Transaction Cost Politics Approach to Policy Making Under Separate Powers199 tbl.8.1, 202–03 tbl.8.2 (1999) (finding that Congress delegates less in the area of taxation than almost any other policy area); James R. Hines Jr. & Kyle D. Logue,Delegating Tax, 114Mich. L. Rev. 235, 248–53 (2015) (explaining that Congress rarely delegates taxing authority to the IRS and Treasury Department, noting that “[i]t is commonly understood that U.S. tax policy is, to a remarkable (and unusual) extent, determined by Congress not only in its broad outlines but also in its details”).

  242. [242]. I.R.C. § 1 (2018) (setting tax rates at specific rates for specific income levels for specific types of income).

  243. [243]. Michael J. Graetz & Deborah H. Schenk, Federal Income Taxation: Principles and Policies 6–12 (6th ed. 2009) (providing a history of the income tax, including battles over the rates and structure).

  244. [244]. Of course, many factors influence Congressional decision-making, but one important factor is the views of their constituents, especially in salient political battles. Gary C. Jacobson & Jamie L. Carson, The Politics of Congressional Elections 261–67 (10th ed. 2019) (finding relationship between constituent beliefs and member votes).

  245. [245]. See generally Dep’t of Transp. Memo, supra note 5 (explaining how the Department of Transportation allocates funds).

  246. [246]. See Cary Coglianese & Christopher S. Yoo, The Bounds of Executive Discretion in the Regulatory State, 164 U. Pa. L. Rev. 1587, 1597 (2016) (describing Congress’s broad delegations of authority to agencies under most statutes); David Epstein & Sharyn O’Halloran,Administrative Procedures, Information, and Agency Discretion, 38Am. J. Pol. Sci. 697, 698–702 (1994)(discussing these reasons for delegation).

  247. [247]. See generally Laura Phillips Sawyer, US Antitrust Law and Policy in Historical Perspective (Harvard Bus. Sch. Working Paper No. 19-110, 2019) (tracing the history of judges’ interpretation of antitrust law).

  248. [248]. Guido Calabresi, A Common Law for the Age of Statutes 32–34 (1982) (tracing the history of judicial interpretation of the Federal Employers Liability Act and common law tort law); Keith N. Hylton, Tort Law: A Modern Perspective 32–47 (2016) (tracing the evolution of tort law and negligence theories).

  249. [249]. David S. Tatel, The Administrative Process and the Rule of Environmental Law, 34 Harv. Env’t L. Rev. 1, 1 (2010) (“As environmental lawyers well know, admin is where the action is.”).

  250. [250]. Sunstein, supra note 15, at 6–21 (discussing the “Triumph of the Technocrats” across the Reagan, Clinton, and Obama administrations); Lisa Heinzerling, Cost-Nothing Analysis: Environmental Economics in the Age of Trump, 30 Colo. Nat. Res., Energy & Env’t L. Rev. 287, 287 (2019) (describing “how, over several decades, cost-benefit analysis came to dominate federal environmental policy”); Landes & Posner, supra note 15, at 387; Elliott Ash, Daniel L. Chen & Suresh Naidu, Ideas Have Consequences: The Impact of Law and Economics on American Justice 8–10 (Mar. 20, 2019) (unpublished manuscript),
    Ideas_Have_Consequences.pdf []) (showing the influence of economics training in the judiciary).

  251. [251]. Lipsey & Lancaster, supra note 22, at 11.

  252. [252]. Steven M. Teles, Kludgeocracy in America, Nat’l Affs. (2013), https://www.national [].

  253. [253]. Liscow, supra note 15, at 1665.

  254. [254]. See generally Council of Econ. Advisers, Expanding Work Requirements in Non-Cash Welfare Programs (2018),
    2018/07/Expanding-Work-Requirements-in-Non-Cash-Welfare-Programs.pdf [
    -KBH3] (giving a history of work requirements for SNAP, Medicaid, and Housing programs since the 1990s).

  255. [255]. Evidence suggests that students can be swayed by economics educations. Robert H. Frank, Thomas Gilovich & Dennis T. Regan, Does Studying Economics Inhibit Cooperation?, 7 J. Econ. Persps. 159, 169 (1993).

  256. [256]. Friedman, supra note 111, at 190–94.

  257. [257]. Mankiw, supra note 11; Polinsky, supra note 9, at 7–8; Cooter & Ulen, supra note 9, at 7.

  258. [258]. See supra Section III.B.

  259. [259]. See Javins v. First Nat’l Realty Corp., 428 F.2d 1071, 1072–73 (D.C. Cir 1970) (establishing the implied warranty of habitability).

  260. [260]. Sunstein, supra note 15, at 40–42 (describing the “easy cases” for efficient policymaking).

  261. [261]. One significant exception to the contracting-around problem is when the price itself is regulated. For example, establishing a modest minimum wage (at least ignoring long-term effects) can shift resources from richer employers to poorer employees. Factors from behavioral economics, such as inattentiveness, could also mitigate the contracting-around problem.

  262. [262]. See, e.g., Paul L. Joskow & Alvin K. Klevorick, A Framework for Analyzing Predatory Pricing Policy, 89 Yale L.J. 213, 222–25 (1979); Frank H. Easterbrook, The Limits of Antitrust, 63 Tex. L. Rev. 1, 3–4 (1984); Alan Devlin & Michael Jacobs, Antitrust Error, 52 Wm. & Mary L. Rev. 75, 97 (2010).

  263. [263]. Kaplow & Shavell, supra note 8, at 27, 30–31; see supra Part II.

  264. [264]. Kaplow & Shavell, supra note 8, at 27, 30–31; see also Kuziemko et al., supra note 226, at 1478–83, 1504–05 (measuring preferences for redistribution).

  265. [265]. See generally Tom R. Tyler, Why People Obey the Law (2006) (describing legitimacy).

  266. [266]. Kaplow & Shavell, supra note 8, at 33–36.

  267. [267]. See generally Dep’t of Transp. Memo, supra note 5 (describing the evaluative procedures of the DOT in assessing passenger time travel).

  268. [268]. Heinzerling, supra note 250, at 288 (describing how standard cost-benefit analysis (“CBA”) is typically an innovation of the executive branch, noting that of all the major environmental statutes, Congress prescribed formal CBA in only one); Robert P. Bartlett III, The Institutional Framework for Cost-Benefit Analysis in Financial Regulation: A Tale of Four Paradigms?, 43 J. Legal Stud. S379, S384–87 (2014) (similar, for financial regulations); Amy Sinden, The Economics of Endangered Species: Why Less is More in the Economic Analysis of Critical Habitat Designations, 28 Harv. Env’t L. Rev. 129, 190–92 (2004) (arguing that, at least the Congresses of the 1970s
    —which were responsible for much of the environmental legislation that has typified Congressional CBA use—not only did not intend to require the formal, highly quantified form of CBA that we now consider standard practice, but in some cases explicitly looked to avoid it).

  269. [269]. See, e.g., Ralph K. Winter, Jr., Poverty, Economic Equality, and the Equal Protection Clause, 1972 Sup. Ct. Rev. 41, 99–102 (1972) (arguing that the Supreme Court should not take distributional concerns into account when applying the Equal Protection Clause). See generally Aditi Bagchi, Distributive Injustice and Private Law, 60 Hastings L.J. 105 (2008) (summarizing the literature and arguing that courts should and often do take distribution into account in private law).

  270. [270]. There is work measuring attitudes for equality in various outcomes. See, e.g., Maureen Cropper, Alan Krupnick & William Raich,Preferences for Equality in Environmental Outcomes 29–33 (Nat’l Bureau of Econ. Rsch., Working Paper No. 22644, 2016). However, I am unaware of any work understanding attitudes about redistributing through specific programs by specific branches of government.

  271. [271]. The causes of reduced growth are myriad, see, e.g., Robert J. Gordon, The Rise and Fall of American Growth: The U.S. Standard of Living Since the Civil War 7–8 (2016) (arguing that “the pace of innovation since 1970 has not been as broad or as deep as that spurred by the inventions of the [last] century”), and the contribution of economic policy is unclear.

  272. [272]. Piketty & Saez, supra note 128, at 11. But see Steven Pinker, Enlightenment Now: The Case for Reason, Science, Humanism, and Progress 322–23 (2018) (arguing that, in many ways, outcomes in the world are improving).

  273. [273]. Uri Friedman, Trust is Collapsing in America, Atlantic (Jan. 21, 2018), https://www.the [

  274. [274]. See generally Pippa Norris, Measuring Populism Worldwide, 26 Party Pol. 697 (2020) (describing populist parties in the United States., Brazil, the Philippines, and elsewhere).

  275. [275]. See also Zachary Liscow & Daniel Markovits, Democraticizing Behavioral Economics, 39 Yale J. on Reg. (forthcoming 2022) (developing one “democratic” approach involving direct engagement with informed, representative members of the public to discern what promotes their well-being).

  276. [276]. HM Treasury, supra note 182, at 40–41 (describing weighting in cost-benefit analysis).

  277. [277]. See, e.g., Emmanuel Saez, Using Elasticities to Derive Optimal Income Tax Rates, 68 Rev. Econ. Stud. 205, 222 (2001).

  278. [278]. See notes 241–43 and accompanying text.

  279. [279]. See text accompanying note 245.


Associate Professor, Yale Law School;

Thanks to Bruce Ackerman, Anne Alstott, Ian Ayres, Joe Bankman, Lily Batchelder, Rachel Bayefsky, Ilan Benshalom, Jake Brooks, Raj Chetty, Conor Clarke, Miranda Fleischer, Brian Galle, David Gamage, Heather Gerken, Jacob Goldin, Henry Hansmann, Alex Hertel-Fernandez, Dan Ho, Paul Kahn, Amy Kapczynski, Noah Kazis, Eugene Kim, Al Klevorick, Lewis Kornhauser, Tony Kronman, Doug Kysar, Leandra Lederman, Daryl Levinson, Yair Listokin, Jon Macey, Daniel Markovits, Alison Morantz, Sam Moyn, Liam Murphy, Ricky Revesz, Diane Ring, Emmanuel Saez, David Schleicher, Scott Shapiro, Dan Shaviro, Steve Sheffrin, Reva Siegel, Manoj Viswanathan, Gui Woolston, Gideon Yaffe, and participants at the Virginia Law School Tax Invitational, University of Michigan Tax Policy Workshop, Duke Law and Economics Workshop, Tulane-Boston College Tax Policy Invitational, NYU Tax and Public Finance Colloquium, Queens Law School, Yale Law School, Indiana University Tax Policy Colloquium, University of British Columbia Tax Policy Colloquium, NYU Law Faculty Workshop, and National Tax Association Annual Meetings for helpful comments and to Audrey Cheng, Zohaib Chida, Benedict Choi, Mia Dana, Kay Dannenmaier, Rebecca Lewis, Joe Liss, Brian McGrail, and Jacob Waggoner for excellent research assistance.